A method for CQI / RI reporting for systems using two-sided CSI AI / ML models

Abstract

A wireless transmit / receive unit (WTRU) using a two-sided model for channel state information (CSI) feedback is configured to detect input / output CSI mismatch between CSI reconstructed at the NW side and CSI estimated at the WTRU side. The WTRU receives configuration information including a CSI mismatch measurement method, a CSI mismatch measurement metric, and one or more thresholds for CSI mismatch detection. The WTRU receives one or more CSI reference signals (CSI-RS) and estimates CSI. The WTRU determines a CSI mismatch event when the mismatch measurement exceeds a first CSI mismatch detection threshold. The WTRU selects a CSI mismatch mitigation method as a function of the I / O CSI mismatch measurement and reports CSI feedback about the received CSI-RS and CSI mismatch information. The CSI mismatch information includes at least one of the I / O CSI mismatch measurement, an indication of the determined CSI mismatch event, or the selected CSI mismatch mitigation method.

Description

[Technical Field]

[0001] The present disclosure relates to a method for CQI / RI reporting for a system using a two-sided CSI AI / ML model. [Background technology]

[0002] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims the benefit of U.S. Provisional Application No. 63 / 445,579, filed February 14, 2023, and U.S. Provisional Application No. 63 / 494,179, filed April 4, 2023, the contents of both of which are incorporated herein by reference.

[0003] In wireless systems, channel state information (CSI) is used between mobile devices and network devices to adapt communications to changing channel conditions. The CSI may include a channel quality index (CQI), a rank indicator (RI), a precoding matrix index (PMI), a Layer 1 (L1) channel measurement (e.g., a reference signal (RS) received power (RSRP) such as L1-RSRP), or a signal-to-interference-and-noise ratio (SINR), a CSI-RS resource indicator (CRI), a synchronization signal physical broadcast channel (SS / PBCH) block resource indicator (SSBRI), a layer indicator (LI), and / or any other measurement quality measured from a configured reference signal (e.g., a CSI-RS or SS / PBCH block or any other reference signal).

[0004] Approaches based on artificial intelligence and machine learning (AI / ML) have the potential to reduce the CSI feedback overhead while maintaining target performance. In contrast to legacy CSI frameworks, AI / ML-based CSI frameworks are two-sided systems, where CSI is generated and possibly compressed on the mobile side, fed back to the base station, e.g., gNB, and restored on the base station side.

[0005] Due to the two-sided nature of AI / ML-based CSI feedback, there may be a mismatch between the precoder calculated on the mobile side and the precoder recovered on the network side. This mismatch can lead to performance degradation because the channel quality indicator (CQI) / rank indicator (RI) reported by the mobile is based on the precoder calculated on the mobile side (X), while the network uses a precoder (X) that may be different.

[0006]

number

[0007] , where:

[0008]

number

[0009] ) because it makes precoding and scheduling decisions based on the precoder mismatch. It would be desirable to determine and report when mismatch between precoders occurs and mitigate performance degradation due to precoder mismatch. For example, strategies are needed to measure, detect, and / or mitigate mismatch between the input and output of a two-sided AI / ML model when mismatch detection is performed on the mobile side and / or when mismatch detection is utilized on the base station side. Additionally, strategies are needed to measure input / output mismatch, detect precoder mismatch, and update the RI and CQI when precoder mismatch is detected for eigenvector-based AI / ML CSI feedback and / or when precoder mismatch is detected for channel matrix-based AI / ML CSI feedback. Summary of the Invention

[0010] In two-sided AI / ML CSI modeling, the content of the output of the CSI generation model in the user equipment (UE), also referred to herein as the wireless transmit / receive unit (WTRU), depends on both the pre-processing used and the AI / ML model used. Therefore, there must be a common understanding between both the WTRU and the base station regarding the pre-processing and AI / ML model used. The output of the CSI generation model can be explicit feedback (i.e., feeding back a compressed version of the channel matrix), implicit feedback (e.g., reusing or modifying the RI / CQI / PMI framework), or a combination of the two. Feeding back a compressed version of the channel matrix may enable optimal feedback reporting quality. However, there may be benefits to sending feedback of the RI and CQI in addition to the channel matrix.

[0011] It may be advantageous to include the RI in the feedback along with the preprocessing and preprocessor selection report. The CQI calculation requires knowledge of interference and may not be derived from the channel matrix alone. For example, the CQI value may be obtained from a combination of channel measurements on a channel measurement resource (CMR) and interference measurements on an interference measurement resource (IMR). Therefore, to ensure that the base station has a complete understanding of the channel conditions at the base station, it is beneficial for the WTRU to report the RI and CQI in addition to the output of the AI / ML encoder.

[0012] In some cases, the output of the base station-side AI / ML model (i.e., the decoder) may not perfectly match the input of the WTRU-side model (e.g., the encoder). In such cases, the reported CQI value may be irrelevant or may be misinterpreted by the base station.

[0013] Although methods have been proposed for detecting CQI mismatches and adjusting the CQI that require the WTRU to have a CSI reconstruction model, this may not always be feasible, for example, when model updates can occur independently. In some cases, the NW must transmit multiple precoded CSI-RSs, each with a different WTRU-specific reconstructed precoder. Alternatively, a demodulation reference signal (DM-RS) may be used instead of the CSI-RS, which may require the allocation of a first physical downlink shared channel (PDSCH), possibly using conservative CQI assumptions.

[0014] CQI mismatch detection in either the WTRU and / or base station can be part of testing / validation of the AI / ML model. However, a small number of occurrences of CQI mismatch should not be a strong enough motivation to determine that the AI / ML model is incompatible. Therefore, it would be beneficial for the WTRU and NW to work together to determine when there is a CQI mismatch.

[0015] In the case of a suspected CQI mismatch, the NW (or WTRU) may provide additional information to the WTRU (or NW) on its reconstructed (or measured) CSI. This may enable it to determine whether there is a mismatch and adjust the CQI. Such additional information may be based, for example, on a metric determined from the difference between the reconstructed / measured CSI and a baseline common CSI assumption.

[0016] Aspects of the disclosed embodiments may address one or more of the above problems by methods and devices for detecting and identifying when there is a mismatch between the AI / ML encoder input of the WTRU and the AI / ML decoder output of the NW.

[0017] In one aspect, a method and device are disclosed for determining and reporting CQI / RI for a system using a two-sided AIML model for CSI feedback, including a method for mitigating potential mismatch between a precoder calculated at the mobile side and a precoder recovered at the NW side. In one example, a WTRU applies adjustment parameters using a two-sided model for CSI feedback that calculates a target CQI based on actual channel measurements.

[0018] In another aspect, a WTRU using a two-sided AI / ML model for CSI feedback determines whether there is a CSI mismatch event, selects a CSI mismatch mitigation method, and reports the mismatch information and the selected mitigation method.

[0019] According to some aspects, the disclosed embodiments may address detecting and mitigating mismatches between inputs and outputs of a two-sided AI / ML CSI model when mismatch detection is performed on the mobile device side. Other aspects may relate to measuring, detecting, and / or mitigating mismatches between inputs and outputs of a two-sided AI / ML model when mismatch detection is performed on the network side, e.g., by a base station or gNB.

[0020] In another aspect, the disclosed embodiments may measure input / output mismatch, detect precoder mismatch, and update RI and CQI when precoder mismatch is detected for eigenvector-based AI / ML CSI feedback. Further aspects of the embodiments relate to updating one or both of RI and CQI when precoder mismatch is detected for channel matrix-based AI / ML CSI feedback.

[0021] Other aspects may be addressed by the WTRU using a two-sided AI / ML model for CSI feedback by determining whether there is a CSI mismatch event and selecting a CSI mismatch mitigation method depending on the input / output CSI mismatch measurements and the available configured mitigation methods.

[0022] In some embodiments, a WTRU using a two-sided AI / ML model for CSI feedback is prompted to send test vectors to the network (NW) for network (NW)-side mismatch detection, and the WTRU applies the configured CSI mismatch mitigation method upon receiving a CSI mismatch indication from the network.

[0023] In another embodiment, a WTRU using a two-sided model for CSI feedback is configured to report an RI / CQI if it determines that an input / output CSI mismatch event has occurred, where the RI / CQI is determined based on one or more received precoded reference signals.

[0024] Further embodiments may relate to using a two-sided model for CSI feedback in which the WTRU is configured to select one or more precoder methods for determining and reporting compressed CSI or RI / CQI when the WTRU determines that an input / output CSI mismatch event has occurred. Additional embodiments are disclosed.

[0025] A more detailed understanding may be had from the following description, given by way of example in conjunction with the accompanying drawings, in which like reference numerals indicate similar elements and in which: [Brief explanation of the drawings]

[0026] [Figure 1A] FIG. 1 is a system diagram illustrating an example communication system in which one or more disclosed embodiments may be implemented. [Figure 1B] 1B is a system diagram illustrating an example wireless transmit / receive unit (WTRU) that may be used within the communication system shown in FIG. 1A, according to an embodiment. [Figure 1C] 1B is a system diagram illustrating an example radio access network (RAN) and an example core network (CN) that may be used within the communication system shown in FIG. 1A, according to an embodiment. [Figure 1D] FIG. 1B is a system diagram illustrating a further exemplary RAN and a further exemplary CN that may be used within the communication system shown in FIG. 1A, according to an embodiment. [Figure 2] FIG. 1 is a functional block diagram illustrating an example of a CSI measurement configuration. [Figure 3] FIG. 1 is a functional block diagram illustrating codebook-based precoding with feedback information. [Figure 4] FIG. 1 illustrates a basic AI / ML framework for CSI feedback. [Figure 5] 10 is a flowchart illustrating a method for a WTRU to measure input / output CSI mismatch in a two-sided AI / ML model, in accordance with an example embodiment. [Figure 6] 1 is a flowchart illustrating a method for NW-side input / output CSI mismatch detection for a two-sided AI / ML model, according to an example embodiment. [Figure 7] FIG. 10 is a signaling diagram illustrating a messaging sequence for updating RI / CQI in eigenvector-based AI / ML CSI feedback according to an example embodiment. [Figure 8] 10 is a flowchart illustrating a method for updating RI / CQI for eigenvector-based AI / ML CSI feedback according to an example embodiment. [Figure 9] FIG. 10 is a signaling diagram illustrating a messaging sequence for updating RI / CQI in AI / ML CSI feedback based on a channel matrix, according to an example embodiment. [Figure 10] 10 is a flowchart illustrating a procedure for selecting a precoder method for full-channel-based AI / ML CSI compression using a two-sided AI / ML model, according to various embodiments. [Figure 11] 10 is a flowchart detailing one example method for a WTRU to measure input / output CSI mismatch for a two-sided AI / ML model. [Figure 12]1 is a flow chart detailing a link adaptation method based on CQI adjustment according to an example embodiment; [Figure 13] 10 is a flow diagram illustrating a method for a WTRU to transmit test vectors to determine CSI mismatch and apply mitigation methods, according to some embodiments. [Figure 14] 10 is a flow diagram illustrating a WTRU method for performing eigenvector-based CSI compression and determining a difference in precoding gain to update a rank indicator (RI) and / or a CQI, according to some embodiments. [Figure 15] 10 is a flow diagram illustrating a WTRU method for reporting RI and / or CQI for CSI compression of the full channel matrix and for a determined precoder method, in accordance with various embodiments. DETAILED DESCRIPTION OF THE INVENTION

[0027] 1A illustrates 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 through the sharing of system resources, including wireless bandwidth. For example, the communication system 100 may utilize one or more channel access methods, such as code division multiple access (CDMA), time division multiple access (TDMA), frequency division multiple access (FDMA), orthogonal FDMA (OFDMA), single-carrier FDMA (SC-FDMA), zero-tail unique word discrete Fourier transform spread OFDM (ZT-UW-DFT-S-OFDM), unique word OFDM (UW-OFDM), resource block filtered OFDM, filter bank multicarrier (FBMC), etc.

[0028] 1A, communications system 100 may include wireless transmit / receive units (WTRUs) 102a, 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, although it will be understood that the disclosed embodiments contemplate any number of WTRUs, base stations, networks, and / or network elements. Each of WTRUs 102a, 102b, 102c, 102d may be any type of device configured to operate and / or communicate in a wireless environment. By way of example, the WTRUs 102a, 102b, 102c, 102d, any of which may be referred to as a base station (STA), may be configured to transmit and / or receive wireless signals and may include user equipment (UE), mobile stations, fixed or mobile subscriber units, contract base units, pagers, 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 wearables, head-mounted displays (HMDs), automobiles, drones, medical devices and applications (e.g., remote surgery), industrial devices and applications (e.g., robots and / or other wireless devices operating in the context of industrial and / or automated processing chains), consumer electronic devices, devices operating on commercial and / or industrial wireless networks, etc. Any of the WTRUs 102a, 102b, 102c, and 102d may be referred to interchangeably as a UE.

[0029] The communications 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 wirelessly interface with at least one of the WTRUs 102a, 102b, 102c, 102d to facilitate access to one or more communications networks, such as the CN 106, the Internet 110, and / or other networks 112. By way of example, the base stations 114a, 114b may be a base transceiver station (BTS), a NodeB, an eNodeB (eNB), a Home Node B, a Home eNodeB, a next generation NodeB such as a gNode B (gNB), a new radio (NR) NodeB, a site controller, an access point (AP), a wireless router, etc. Although the base stations 114a, 114b are each illustrated as a single element, it will be understood that the base stations 114a, 114b may include any number of interconnected base stations and / or network elements.

[0030] 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), relay nodes, etc. The base station 114a and / or base station 114b may be configured to transmit and / or receive wireless signals on one or more carrier frequencies, sometimes referred to as a cell (not shown). These frequencies may be in the licensed spectrum, the unlicensed spectrum, or a combination of the licensed and unlicensed spectrum. A cell may provide coverage for wireless services in a particular geographic area, which may be relatively constant or may change over time. A cell may be further divided into cell sectors. For example, the 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, i.e., one transceiver for each sector of the cell. In an embodiment, the base station 114a may utilize multiple-input multiple-output (MIMO) technology and may utilize multiple transceivers for each sector of the cell. For example, beamforming may be used to transmit and / or receive signals in desired spatial directions.

[0031] 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 communications 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).

[0032] More specifically, as mentioned above, the communications system 100 may be a multiple-access system and may utilize one or more channel access schemes, such as CDMA, TDMA, FDMA, OFDMA, SC-FDMA, etc. For example, the base station 114a and the WTRUs 102a, 102b, 102c in the RAN 104 may implement a radio technology such as Universal Mobile Telecommunications System (UMTS) Terrestrial Radio Access (UTRA), which may establish the air interface 116 using Wideband CDMA (WCDMA). 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).

[0033] 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 LTE-Advanced (LTE-A) and / or LTE-Advanced Pro (LTE-A Pro).

[0034] 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.

[0035] 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 jointly implement LTE radio access and NR radio access, e.g., using the principle of dual connectivity (DC). Thus, the air interface utilized by the WTRUs 102a, 102b, 102c may be characterized by multiple types of radio access technologies and / or transmissions transmitted to / from multiple types of base stations (e.g., eNBs and gNBs).

[0036] In other embodiments, the base station 114a and the WTRUs 102a, 102b, 102c may implement a wireless technology such as IEEE 802.11 (i.e., Wireless Fidelity (WiFi)), IEEE 802.16 (i.e., Worldwide Interoperability for Microwave Access (WiMAX)), CDMA2000, CDMA2000 1X, CDMA2000 EV-DO, Interim Standard 2000 (IS-2000), Interim Standard 95 (IS-95), Interim Standard 856 (IS-856), Global System for Mobile communications (GSM), Enhanced Data rates for GSM Evolution (EDGE), GSM EDGE (GERAN), or the like.

[0037] 1A may be a wireless router, Home Node B, Home eNode B, or access point and may utilize any suitable RAT for facilitating wireless connectivity in a localized area, such as a business, residence, automobile, campus, industrial facility, air corridor (e.g., for use by drones), road, etc. In one embodiment, the base station 114b and the WTRUs 102c, 102d may implement a radio technology such as IEEE 802.11 to establish a wireless local area network (WLAN). In an embodiment, the base station 114b and the WTRUs 102c, 102d may implement a radio technology such as IEEE 802.15 to establish a wireless personal area network (WPAN). In yet another embodiment, the base station 114b and the WTRUs 102c, 102d may utilize a cellular-based RAT (e.g., WCDMA, CDMA2000, GSM, LTE, LTE-A, LTE-A Pro, NR, etc.) to establish a picocell or femtocell. 1A, the base station 114b may have a direct connection to the Internet 110. Therefore, the base station 114b may not be required to access the Internet 110 via the CN 106.

[0038] The RAN 104 may be in communication with the CN 106, which may be any type of network configured to provide voice, data, application, and / or Voice over Internet Protocol (VoIP) services to one or more of the WTRUs 102a, 102b, 102c, 102d. The data may have various quality of service (QoS) requirements, such as different throughput requirements, latency requirements, error resilience requirements, reliability requirements, data throughput requirements, mobility requirements, etc. The CN 106 may provide call control, billing services, mobile location-based services, prepaid calling, Internet connectivity, video distribution, etc., and / or perform high-level security functions such as user authentication. Although not shown in FIG. 1A , it will be understood that the RAN 104 and / or CN 106 may be in direct or indirect communication with other RANs that utilize the same RAT as the RAN 104 or a different RAT. For example, in addition to being connected to the RAN 104, which may utilize NR radio technology, the CN 106 may also be in communication with another RAN (not shown) that utilizes GSM, UMTS, CDMA2000, WiMAX, E-UTRA, or WiFi radio technology.

[0039] 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 providing 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 Transmission Control Protocol (TCP), User Datagram Protocol (UDP), and / or Internet Protocol (IP) of 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 network 112 may include another CN connected to one or more RANs, which may utilize the same RAT as the RAN 104 or a different RAT.

[0040] Some or all of the WTRUs 102a, 102b, 102c, 102d in the communications system 100 may include multi-mode capabilities (e.g., the WTRUs 102a, 102b, 102c, 102d may include multiple transceivers for communicating with different wireless networks over different wireless links). For example, the WTRU 102c shown in FIG. 1A may be configured to communicate with a base station 114a that may utilize cellular-based wireless technology and with a base station 114b that may utilize IEEE 802.11 wireless technology.

[0041] 1B is a system diagram illustrating an example WTRU 102. As shown in FIG. 1B, the WTRU 102 may include, among other things, a processor 118, a transceiver 120, a transmit / receive element 122, a speaker / microphone 124, a keypad 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 other peripherals 138. It will be understood that the WTRU 102 may include any sub-combination of the above-described elements while remaining consistent with an embodiment.

[0042] The processor 118 may be a general-purpose processor, a special-purpose processor, a conventional processor, a digital signal processor (DSP), multiple microprocessors, one or more microprocessors associated with a DSP core, a controller, a microcontroller, an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), 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. While FIG. 1B illustrates the processor 118 and the transceiver 120 as separate components, it will be understood that the processor 118 and the transceiver 120 may be integrated together in an electronic package or chip.

[0043] The transmit / receive element 122 may be configured to transmit signals to or receive signals from a base station (e.g., base station 114a) over the air interface 116. For example, in one embodiment, the transmit / receive element 122 may be an antenna configured to transmit and / or receive RF signals. In an embodiment, the transmit / receive element 122 may be an emitter / detector configured to transmit and / or receive IR, UV, or visible light signals, for example. In yet another embodiment, the transmit / receive element 122 may be configured to transmit and / or receive both RF and light signals. It will be understood that the transmit / receive element 122 may be configured to transmit and / or receive any combination of wireless signals.

[0044] 1B as a single element, the WTRU 102 may include any number of transmit / receive elements 122. More specifically, the WTRU 102 may utilize 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.

[0045] The transceiver 120 may be configured to modulate signals to be transmitted by the transmit / receive element 122 and demodulate signals received by the transmit / receive element 122. As mentioned 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, such as, for example, NR and IEEE 802.11.

[0046] The processor 118 of the WTRU 102 may be coupled to and may receive user input data from a speaker / microphone 124, a keypad 126, and / or a display / touchpad 128 (e.g., a liquid crystal display (LCD) display unit or an organic light emitting diode (OLED) display unit). The processor 118 may also output user data to the speaker / microphone 124, the keypad 126, and / or the display / touchpad 128. Additionally, the processor 118 may access information from and store data in any type of suitable memory, such as non-removable memory 130 and / or removable memory 132. The non-removable memory 130 may include random access memory (RAM), 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, etc. In other embodiments, the processor 118 may access information from, and store data in, memory that is not physically located on the WTRU 102, such as on a server or home computer (not shown).

[0047] The processor 118 may receive power from the power source 134 and may be configured to distribute and / or control the power to 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 batteries (e.g., nickel-cadmium (NiCd), nickel-zinc (NiZn), nickel-metal hydride (NiMH), lithium-ion (Li-ion), etc.), solar cells, fuel cells, etc.

[0048] The processor 118 may also be coupled to a 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 instead of, information from the GPS chipset 136, the WTRU 102 may receive location information from base stations (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 will be understood that the WTRU 102 may acquire location information by way of any suitable location determination method while remaining consistent with an embodiment.

[0049] The processor 118 may further be coupled to other 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, the peripherals 138 may include an accelerometer, an electronic compass, a satellite transceiver, a digital camera (for photos and / or videos), a universal serial bus (USB) port, a vibration device, a television transceiver, a hands-free headset, a Bluetooth module, a frequency modulation (FM) radio unit, a digital music player, a media player, a video game player module, an internet browser, a virtual reality and / or augmented reality (VR / AR) device, an activity tracker, etc. The peripherals 138 may include one or more sensors. The sensors 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.

[0050] The WTRU 102 may include a full-duplex radio that is associated with a particular subframe for both UL (e.g., for transmission) and DL (e.g., for reception) transmission and reception of some or all of the signals (e.g., which may be contemporaneous and / or simultaneous). The full-duplex radio may include an interference management unit for reducing and / or substantially eliminating self-interference, either via hardware (e.g., a choke) or via signal processing via a processor (e.g., via a separate processor (not shown) or via processor 118). In an embodiment, the WTRU 102 may include a half-duplex radio that is associated with a particular subframe for either UL (e.g., for transmission) or DL ​​(e.g., for reception) transmission and reception of some or all of the signals (e.g., which may be contemporaneous and / or simultaneous).

[0051] 1C is a system diagram illustrating the RAN 104 and the CN 106, according to an embodiment. As mentioned above, the RAN 104 may utilize 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.

[0052] While the RAN 104 may include eNode-Bs 160a, 160b, and 160c, it will be understood that the RAN 104 may include any number of eNode-Bs while remaining consistent with an embodiment. The eNode-Bs 160a, 160b, and 160c may each include one or more transceivers for communicating with the WTRUs 102a, 102b, and 102c over the air interface 116. In one embodiment, the eNode-Bs 160a, 160b, and 160c may implement MIMO technology. Thus, the eNode-B 160a may use multiple antennas to transmit wireless signals to and / or receive wireless signals from the WTRU 102a, for example.

[0053] Each of the eNode-Bs 160a, 160b, 160c may be associated with a particular cell (not shown) and may be configured to handle radio resource management decisions, handover decisions, scheduling of users in the UL and / or DL, etc. As shown in FIG. 1C, the eNode-Bs 160a, 160b, 160c may communicate with one another via an X2 interface.

[0054] 1C may include a mobility management entity (MME) 162, a serving gateway (SGW) 164, and a packet data network (PDN) gateway (PGW) 166. Although the above-mentioned elements are illustrated as part of the CN 106, it will be understood that any of these elements may be owned and / or operated by an entity other than the CN operator.

[0055] 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 act as a control node. For example, the MME 162 may be responsible for authenticating users of the WTRUs 102a, 102b, 102c, bearer activation / deactivation, selecting a particular serving gateway during initial attach of the WTRUs 102a, 102b, 102c, etc. 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.

[0056] 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 handovers between eNode Bs, triggering paging when DL data is available for the WTRUs 102a, 102b, 102c, managing and storing the context of the WTRUs 102a, 102b, 102c, etc.

[0057] The SGW 164 may be connected to a 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.

[0058] The CN 106 may facilitate communication 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, to facilitate communication between the WTRUs 102a, 102b, 102c and traditional landline communication 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.

[0059] Although the WTRUs are described in Figures 1A-1D as wireless terminals, it is contemplated that in some representative embodiments, such terminals may use a wired communication interface (e.g., temporary or permanent) with the communication network.

[0060] In an exemplary embodiment, the other network 112 may be a WLAN.

[0061] A WLAN in infrastructure basic service set (BSS) mode may have an access point (AP) for the BSS and one or more stations (STAs) associated with the AP. The AP may have access to or interface with a distribution system (DS) or another type of wired / wireless network that carries traffic to and / or from the BSS. Traffic to a STA originating from outside the BSS may arrive and be delivered to the STA through the AP. Traffic originating from a STA toward a destination outside the BSS may be transmitted to the AP to be delivered to the respective destination. Traffic between STAs within a BSS may be transmitted through the AP; e.g., a source STA may transmit traffic to the AP, and the AP may deliver the traffic to the destination STA. Traffic between STAs within a BSS may be considered and / or referred to as peer-to-peer traffic. Peer-to-peer traffic may be transmitted between (e.g., directly between) a source STA and a destination STA using direct link setup (DLS). In some representative embodiments, the DLS may use 802.11e DLS or 802.11z Tunneled DLS (TDLS). A WLAN using an Independent BSS (IBSS) mode may not have an AP, and STAs within or using the IBSS (e.g., all of the STAs) may communicate directly with each other. The IBSS communication mode is sometimes referred to herein as an "ad hoc" communication mode.

[0062] When using the 802.11ac infrastructure mode of operation or a similar mode of operation, an AP may transmit beacons on a fixed channel, such as a primary channel. The primary channel may have a fixed width (e.g., a 20 MHz wide bandwidth) or a dynamically configured width. The primary channel may be the operating channel of the BSS and may be used by STAs to establish a connection with the AP. In some representative embodiments, Carrier Sense Multiple Access with Collision Avoidance (CSMA / CA) may be implemented, for example, in an 802.11 system. In CSMA / CA, STAs (e.g., all STAs), including the AP, may sense the primary channel. If the primary channel is sensed / detected and / or determined to be busy by a particular STA, that particular STA may yield. One STA (e.g., only one station) may transmit at any given time within a given BSS.

[0063] A high-throughput (HT) STA may use, for example, a 40 MHz wide channel for communication via a combination of a primary 20 MHz channel and adjacent or non-adjacent 20 MHz channels to form a 40 MHz wide channel.

[0064] A very high throughput (VHT) STA may support channels with widths of 20 MHz, 40 MHz, 80 MHz, and / or 160 MHz. A 40 MHz and / or 80 MHz channel may be formed by combining contiguous 20 MHz channels. A 160 MHz channel may be formed by combining eight contiguous 20 MHz channels or by combining two non-contiguous 80 MHz channels, which may be referred to as an 80+80 configuration. In the 80+80 configuration, data after channel coding may be passed through a segment parser that may split the data into two streams. Inverse fast Fourier transform (IFFT) processing and time-domain processing may be performed separately on each stream. The streams may be mapped to two 80 MHz channels, and the data may be transmitted by the transmitting STA. At the receiver of the receiving STA, the operations described above for the 80+80 configuration may be reversed, and the combined data may be transmitted to the medium access control (MAC).

[0065] Sub-1 GHz operating modes are supported by 802.11af and 802.11ah. Channel operating bandwidths and carriers are reduced in 802.11af and 802.11ah relative to those used in 802.11n and 802.11ac. 802.11af supports 5 MHz, 10 MHz, and 20 MHz bandwidths in the TV White Space (TVWS) spectrum, while 802.11ah supports 1 MHz, 2 MHz, 4 MHz, 8 MHz, and 16 MHz using non-TVWS spectrum. According to a representative embodiment, 802.11ah may support meter-type control / MTC, such as machine-type communication (MTC) devices in macro coverage areas. MTC devices may have limited capabilities, for example, support for (e.g., only) some and / or limited bandwidths. MTC devices may include batteries with battery life above a certain threshold (e.g., to maintain very long battery life).

[0066] WLAN systems that can support multiple channels and channel bandwidths, such as 802.11n, 802.11ac, 802.11af, and 802.11ah, include a channel that can be designated as a primary channel. The primary channel can have a bandwidth equal to the largest common operating bandwidth supported by all STAs in the BSS. The bandwidth of the primary channel can be set and / or limited by one of all STAs operating in the BSS that supports the minimum bandwidth operating mode. In an 802.11ah example, the primary channel can be 1 MHz wide for STAs (e.g., MTC-type devices) that support (e.g., only) 1 MHz mode, even if the AP and other STAs in the BSS support 2 MHz, 4 MHz, 8 MHz, 16 MHz, and / or other channel bandwidth operating modes. Carrier sensing and / or network allocation vector (NAV) setting can be contingent on the status of the primary channel. If the primary channel is busy, for example, due to a STA (that only supports 1 MHz mode of operation) transmitting to the AP, all available frequency bands may be considered busy even if the majority of the available frequency bands remain idle.

[0067] In the United States, the available frequency band that can be used by 802.11ah is 902 MHz to 928 MHz. In South Korea, the available frequency band is 917.5 MHz to 923.5 MHz. In Japan, the available frequency band is 916.5 MHz to 927.5 MHz. The total available bandwidth for 802.11ah is 6 MHz to 26 MHz depending on the country code.

[0068] 1D is a system diagram illustrating the RAN 104 and the CN 106 in accordance with an embodiment. As mentioned above, the RAN 104 may utilize 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.

[0069] While the RAN 104 may include gNBs 180a, 180b, and 180c, it will be understood that the RAN 104 may include any number of gNBs while remaining consistent with an embodiment. The gNBs 180a, 180b, and 180c may each include one or more transceivers for communicating with the WTRUs 102a, 102b, and 102c over the air interface 116. In one embodiment, the gNBs 180a, 180b, and 180c may implement MIMO technology. For example, the gNBs 180a and 180b may utilize beamforming to transmit signals to and / or receive signals from the gNBs 180a, 180b, and 180c. Thus, the gNB 180a may, for example, transmit wireless signals to and / or receive wireless signals from the WTRU 102a using multiple antennas. In an embodiment, the gNBs 180a, 180b, and 180c may implement carrier aggregation technology. For example, the gNB 180a may transmit multiple component carriers to the WTRU 102a (not shown). A subset of these component carriers may be on unlicensed spectrum, while the remaining component carriers may be on licensed spectrum. In an embodiment, the gNBs 180a, 180b, 180c may implement Coordinated Multi-Point (CoMP) technology. For example, the WTRU 102a may receive coordinated transmissions from the gNBs 180a and 180b (and / or 180c).

[0070] The WTRUs 102a, 102b, 102c may communicate with the gNBs 180a, 180b, 180c using transmissions associated with scalable numerologies. For example, the OFDM symbol spacing and / or OFDM subcarrier spacing may be different 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., including a variable number of OFDM symbols and / or variable lengths of absolute time duration).

[0071] 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 another RAN (e.g., eNode-Bs 160a, 160b, 160c, etc.). In a standalone configuration, the WTRUs 102a, 102b, 102c may utilize one or more of the gNBs 180a, 180b, 180c as mobility anchor points. In a standalone configuration, the WTRUs 102a, 102b, 102c may communicate with the gNBs 180a, 180b, 180c using signals in unlicensed spectrum. In a non-standalone configuration, the WTRUs 102a, 102b, 102c may communicate / connect with a gNB 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 DC principles 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 act 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.

[0072] Each of the gNBs 180a, 180b, 180c may not be associated with a particular cell (not shown) and may be configured to handle radio resource management decisions, handover decisions, scheduling of users in the UL and / or DL, support for network slicing, DC, interconnection between NR and E-UTRA, routing of user plane data towards user plane functions (UPFs) 184a, 184b, routing of control plane information towards access and mobility management functions (AMFs) 182a, 182b, etc. As shown in FIG. 1D , the gNBs 180a, 180b, 180c may communicate with one another over an Xn interface.

[0073] 1D may include at least one AMF 182a, 182b, at least one UPF 184a, 184b, at least one Session Management Function (SMF) 183a, 183b, and possibly a Data Network (DN) 185a, 185b. While the above-mentioned elements are illustrated as part of the CN 106, it will be understood that any of these elements may be owned and / or operated by an entity other than the CN operator.

[0074] The AMF 182a, 182b may be connected to one or more of the gNBs 180a, 180b, 180c in the RAN 104 via an N2 interface and may act as a control node. For example, the AMF 182a, 182b may be responsible for authenticating users of the WTRUs 102a, 102b, 102c, supporting network slicing (e.g., handling different protocol data unit (PDU) sessions with different requirements), selecting a particular SMF 183a, 183b, managing registration areas, terminating non-access stratum (NAS) signaling, mobility management, etc. Network slicing may be used by the AMF 182a, 182b to customize CN support for the WTRUs 102a, 102b, 102c based on the type of service utilized by the WTRUs 102a, 102b, 102c. For example, different network slices may be established for different use cases, such as services relying on Ultra-Reliable Low-Latency (URLLC) access, services relying on enhanced Massive Mobile Broadband (eMBB) access, services for MTC access, etc. The AMFs 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-A Pro, and / or non-3GPP access technologies, such as WiFi.

[0075] The SMFs 183a, 183b may be connected to the AMFs 182a, 182b in the CN 106 via an N11 interface. The SMFs 183a, 183b may also be connected to the UPFs 184a, 184b in the CN 106 via an N4 interface. The SMFs 183a, 183b may select and control the UPFs 184a, 184b and configure the routing of traffic through the UPFs 184a, 184b. The SMFs 183a, 183b may perform other functions such as managing and allocating UE IP addresses, managing PDU sessions, controlling policy management and QoS, providing DL data notification, etc. The PDU session type may be IP-based, non-IP-based, Ethernet-based, etc.

[0076] The UPFs 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 UPFs 184a, 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.

[0077] The CN 106 may facilitate communication 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 DNs 185a, 185b through the UPFs 184a, 184b via an N3 interface to the UPFs 184a, 184b and an N6 interface between the UPFs 184a, 184b and the DNs 185a, 185b.

[0078] 1A-1D and the corresponding description thereof, one or more, or all, of the functions described herein with respect to one or more of the WTRUs 102a-d, base stations 114a-b, eNode-Bs 160a-c, MME 162, SGW 164, PGW 166, gNBs 180a-c, AMFs 182a-b, UPFs 184a-b, SMFs 183a-b, DNs 185a-b, and / or any other devices described herein may be performed by one or more emulation devices (not shown). The emulation devices may be one or more devices configured to emulate one or more, or all, of the functions described herein. For example, the emulation devices may be used to test other devices and / or to simulate network and / or WTRU functionality.

[0079] The emulation device may be designed to perform one or more tests of other devices in a laboratory environment and / or in an operator network environment. For example, one or more emulation devices may perform one or more, or all of the functions while fully or partially implemented and / or deployed as part of a wired and / or wireless communications network to test other devices in the communications network. One or more emulation devices may perform one or more, or all of the functions while temporarily implemented / deployed as part of a wired and / or wireless communications network. The emulation device may be directly coupled to another device for the purpose of testing and / or performing tests using over-the-air wireless communications.

[0080] The one or more emulation devices may perform one or more of the functions, including all of the functions, without being implemented / deployed as part of a wired and / or wireless communications network. For example, the emulation devices may be utilized in a test lab and / or in a test scenario in an undeployed (e.g., experimental) wired and / or wireless communications network to perform testing of one or more components. The one or more emulation devices may be test equipment. Direct RF coupling and / or wireless communication via RF circuitry (e.g., which may include one or more antennas) may be used by the emulation devices to transmit and / or receive data.

[0081] For CSI feedback based on AI / ML, embodiments include methods for determining and reporting CQI / RI for systems using a two-sided AIML model for CSI feedback, including methods for mitigating potential mismatch between the precoder calculated at the WTRU side and the precoder recovered at the network (NW) side.

[0082] As previously mentioned, the channel state information (CSI) may include at least one of a channel quality index (CQI), a rank indicator (RI), a precoding matrix index (PMI), an L1 channel measurement (e.g., a reference signal received power (RSRP) such as L1-RSRP or a signal-to-interference-and-noise ratio (SINR), a CSI-RS resource indicator (CRI), an SS / PBCH block resource indicator (SSBRI), a layer indicator (LI), and / or any other measurement quality measured by the WTRU from a configured reference signal (e.g., a CSI-RS or SS / PBCH block or any other reference signal).

[0083] An example CSI reporting framework is described herein. The WTRU may be configured to report CSI over an uplink (UL) control channel, such as a physical uplink control channel (PUCCH), or by gNB request for a physical uplink shared channel (PUSCH) grant. Depending on the configuration, the CSI reference signal (CSI-RS) can cover the entire bandwidth of the bandwidth portion (BWP) or only a portion of it. Within the CSI-RS bandwidth, CSI-RS may be configured in each physical resource block (PRB) or every other PRB. In the time domain, 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 resources can be (de)activated by a medium access control (MAC) control element (CS), and the WTRU reports associated measurements only when the resources are activated. In aperiodic CSI-RS, the WTRU is triggered to report the measured CSI-RS on the PUSCH by a request in the downlink control information (DCI). Periodic reports are carried on the PUCCH, while semi-persistent reports can be carried on either the PUCCH or the PUSCH. The reported CSI can be used by the scheduler when allocating resource blocks, possibly based on the time-frequency selectivity of the channel, when determining precoding matrices, beams, transmission modes, and / or when selecting an appropriate modulation and coding scheme (MCS). The reliability, accuracy, and timeliness of WTRU CSI reporting can be important to meet Ultra-Reliable Low-Latency Communications (URLLC) service requirements.

[0084] 2 , an example configuration 200 for a CSI measurement configuration is shown. A WTRU may be configured with a CSI measurement configuration, which may include one or more CSI reporting configurations 205, 208, one or more resource configurations 210, 212, 214, and / or one or more links 230, 232, 234, 236 between the one or more CSI reporting configurations 205, 208 and the one or more resource configurations 210, 212, 214. In the CSI measurement configuration, one or more of the following configuration parameters may be provided:

[0085] (1) N≧1 CSI reporting configurations, M≧1 resource configurations, and CSI measurement configurations that link the N CSI reporting configurations and the M resource configurations.

[0086] (2) CSI reporting configuration, including one or more of the following: time domain behavior, i.e., aperiodic or periodic / semi-persistent, frequency granularity for at least the precoding matrix index (PMI) and CQI, CSI reporting type (e.g., PMI, CQI, RI, CRI, etc.), and / or, if PMI is reported, PMI type (Type I or Type II) and codebook configuration.

[0087] (3) Time domain behavior: aperiodic or periodic / semi-persistent, RS type (e.g., for channel measurement or interference measurement), and / or resource configuration including one or more of S≧1 resource sets, where each resource set may contain Ks resources.

[0088] (4) The CSI measurement configuration includes one or more of a CSI reporting configuration, a resource configuration, and / or a reference transmission scheme configuration for CQI; and / or

[0089] (5) For CSI reporting for a component carrier (CC), one or more frequency granularities may be supported, including wideband CSI, partial band CSI, and subband CSI.

[0090] 3, a basic example of codebook-based precoding with feedback information 300 is shown. The feedback information may include a precoding matrix index (PMI), which may be referred to as a codeword index in the codebook as shown in the figure.

[0091] As shown in Figure 3, the codebook includes a set of precoding vectors / matrices for each rank and the number of antenna ports, and each precoding vector / matrix has its own unique index, so the receiver 305 can feed back 308 the preferred precoding vector / matrix index to the transmitter 310. Codebook-based precoding may have performance degradation compared to non-codebook-based precoding due to its finite number of precoding vectors / matrices. However, a major advantage of codebook-based precoding may be lower control signaling / feedback overhead. Table 1 below shows an example codebook for 2Tx.

[0092] [Table 1]

[0093] Artificial intelligence (AI) can be broadly defined as behavior exhibited by machines that can mimic cognitive functions, for example, to sense, think, adapt, and act.

[0094] Machine learning (ML) may refer to a type of algorithm that solves problems based on learning through experience (“data”) without being explicitly programmed (“constructing a set of rules”). Machine learning may be considered a subset of AI. Different machine learning paradigms may be envisioned based on the nature of the data or feedback available to the learning algorithm. For example, a supervised learning approach may involve learning a function that maps inputs to outputs based on labeled training examples, where each training example may be a pair consisting of an input and a corresponding output. For example, an unsupervised learning approach may involve detecting patterns in existing unlabeled data. For example, a reinforcement learning approach may involve performing a sequence of actions in an environment to maximize a cumulative reward. In some strategies, it is possible to apply machine learning algorithms using a combination or interpolation of the above-mentioned approaches. For example, a semi-supervised learning approach may use a combination of a small amount of labeled data and a large amount of unlabeled data during training. In this regard, semi-supervised learning lies between unsupervised learning (without labeled training data) and supervised learning (with only labeled training data).

[0095] Deep learning refers to a category of machine learning algorithms that utilize artificial neural networks (specifically DNNs), loosely inspired by biological systems. Deep neural networks (DNNs) are a special class of machine learning models inspired by the human brain, in which inputs are linearly transformed and passed through nonlinear activation functions multiple times. DNNs typically consist of multiple layers, each consisting of a linear transformation and a given nonlinear activation function. DNNs can be trained using training data via the backpropagation algorithm. Recently, DNNs have demonstrated state-of-the-art performance in various domains, such as speech, vision, and natural language, as well as in various supervised, unsupervised, and semi-supervised machine learning settings. The term AIML-based methods / processes can refer to achieving behavior and / or adapting to requirements through data-based learning without an explicit configuration of a sequence of behavioral steps. Such methods can enable learning complex behaviors that may be difficult to specify and / or implement using legacy methods.

[0096] 4, an AI / ML-based CSI feedback framework 400 may use an autoencoder (AE), also called a precoder, for CSI compression. This is a two-sided system, where estimated CSI is compressed by an encoder 405 at the WTRU side and fed back to the gNB (410), and the compressed CSI is restored by a decoder 415 at the gNB.

[0097] Machine learning-based approaches (e.g., autoencoder-AE) have the potential to reduce the overhead of CSI feedback while maintaining the target performance. In contrast to legacy CSI frameworks, AI / ML-based CSI frameworks are two-sided systems, where CSI is generated and possibly compressed on the WTRU side, fed back to the gNB, and restored on the gNB side.

[0098] Due to the bilateral nature of the AI / ML-based CSI feedback, the precoder (X) calculated at the WTRU side and the precoder (

[0099]

number

[0100] ) which may lead to performance degradation because the CQI / RI reported by the WTRU is based on the precoder (X) calculated at the WTRU side, while the NW uses a possibly different precoder (

[0101]

number

[0102] , where:

[0103]

number

[0104] ) is the basis for making precoding and scheduling decisions.

[0105] Embodiments disclosed herein may determine and report when a mismatch occurs between precoders and mitigate performance degradation due to precoder mismatch. In various embodiments, methods and devices are disclosed for measuring, detecting, and mitigating mismatch between the input and output of a two-sided AI / ML model when mismatch detection is performed at the WTRU side. Other embodiments relate to measuring, detecting, and mitigating mismatch between the input and output of a two-sided AI / ML model when mismatch detection is utilized at the gNB. Further embodiments disclosed below relate to measuring input / output mismatch, detecting precoder mismatch, and updating RI and CQI when precoder mismatch is detected for eigenvector-based AI / ML CSI feedback. Yet further embodiments disclose methods for updating RI and CQI when precoder mismatch is detected for channel matrix-based AI / ML CSI feedback.

[0106] A method for a WTRU to measure input / output CSI mismatch for a two-sided AI / ML model may generally include WTRU configuration, WTRU measurement, reporting, and WTRU mitigation, as described in further detail below. In one exemplary embodiment, a WTRU using a two-sided model for CSI feedback is configured to measure the mismatch between the CSI reconstructed at the NW side and the CSI estimated at the WTRU side (e.g., input / output CSI mismatch), and the configuration includes a CSI mismatch measurement method, a CSI mismatch metric, and one or more thresholds for CSI mismatch detection.

[0107] When configured, the WTRU receives the CSI-RS and estimates the CSI. The WTRU determines an input / output CSI mismatch measurement value based on the CSI mismatch metric and / or the estimated CSI and / or the configured CSI mismatch measurement method. The WTRU then determines whether there is a CSI mismatch event based on the input / output CSI mismatch measurement value and a first configured CSI mismatch detection threshold.

[0108] When the WTRU detects a CSI mismatch event, the WTRU may select a configured CSI mismatch mitigation method depending on the incoming / outgoing CSI mismatch measurement. For example, in one embodiment, if CSI feedback reporting is configured to use the data channel (PUSCH), the WTRU reduces the CSI feedback compression ratio when the incoming / outgoing CSI mismatch measurement exceeds a second configured CSI mismatch detection threshold. Alternatively, if CSI feedback reporting is configured to use the control channel (PUCCH), the WTRU requests to switch to data channel reporting, for example, when the CSI mismatch measurement is lower than the second configured CSI mismatch detection threshold. In one embodiment, the WTRU then reports CSI feedback and CSI mismatch information, including the incoming / outgoing CSI mismatch measurement, or an indication that a CSI mismatch event has occurred, and / or a preferred CSI mismatch mitigation.

[0109] WTRU procedures and reporting for NW-side input / output CSI mismatch detection for a two-sided AI / ML model may generally include WTRU configuration for measurement and reporting. In these embodiments, a WTRU using a two-sided model for CSI feedback is configured to support NW-side measurement and detection of mismatch between NW-side reconstructed CSI and WTRU-side estimated CSI (input / output CSI mismatch).

[0110] In one exemplary embodiment, the WTRU configuration includes a test vector type (standalone, partial input, input to preprocessing, or input to AI / ML encoder), a set of test vectors or preconfigured patterns, test vector selection criteria, a metric to monitor, and one or more CSI mismatch mitigation methods. The WTRU is triggered to transmit the test vectors to the NW by at least one of a time trigger (e.g., based on a preconfigured period and offset) and / or an event trigger (e.g., based on the monitored metric exceeding a configured threshold).

[0111] The WTRU selects one or more test vector types and / or test vectors based on the test vector selection criteria and the monitored metrics, and transmits the selected test vectors to the NW when prompted. If the WTRU receives a CSI mismatch indication from the NW, the WTRU selects and / or applies a configured CSI mismatch mitigation method, for example, the WTRU reduces the CSI compression rate, or the WTRU switches to a different AI / ML encoder model, or the WTRU switches or disables preprocessing, or the WTRU requests switching to PUSCH for CSI feedback reporting (e.g., if PUCCH was used). The WTRU then transmits CSI feedback to the NW based on the selected CSI mismatch mitigation method.

[0112] A WTRU method for updating the RI / CQI for eigenvector-based AI / ML CSI feedback using a two-sided AI / ML model may generally include a WTRU using a two-sided model for CSI feedback being configured to report the RI / CQI if it determines that an input / output CSI mismatch event has occurred. In one embodiment, this configuration may include parameters for performing eigenvector (EV)-based CSI compression, one or more thresholds (e.g., precoding gain thresholds) for determining an input / output CSI mismatch event, and a reporting configuration for the compressed CSI feedback.

[0113] The WTRU receives the CSI-RS, determines CSI (including a first rank indicator (RI) and a channel quality indicator (CQI)), calculates an original precoding gain, and performs EV-based CSI compression. The WTRU reports the first RI and CQI, and reports compressed CSI (e.g., a first precoder or precoder matrix, or an indication thereof, associated with the first determined RI and CQI). The WTRU then receives a reference signal (RS) precoded with a second precoder (e.g., the second precoder is determined in the gNB and may be different from the first precoder). The WTRU measures an effective precoding gain based on the received RS precoded with the second precoder, where the effective precoding gain is the gain of the precoded channel. The WTRU determines a second RI and CQI when the difference between the measured effective precoding gain and the original precoding gain exceeds a configured precoding gain threshold, and reports the second RI and CQI.

[0114] In other embodiments, methods for selecting a precoder method for AI / ML CSI compression based on the entire channel using a two-sided AI / ML model are disclosed. In these embodiments, a WTRU using a two-sided model for CSI feedback is configured to select one or more precoder methods for determining and reporting compressed CSI or RI / CQI. One example configuration may include a set of precoder methods (e.g., singular value decomposition (SVD), zero-forcing (ZF)) for determining a precoder, one or more precoder method selection thresholds, and a reporting configuration for compressed CSI feedback and RI / CQI. The WTRU receives the CSI-RS and performs CSI compression of the entire channel matrix. The WTRU then determines one or more precoder methods depending on the reporting configuration, measured channel conditions, and / or one or more precoder method selection thresholds. In various exemplary embodiments, determining the precoder method may include selecting the precoder method that results in the highest CQI or RI, selecting the precoder method based on measurements, and / or selecting the precoder method based on a determined input / output CSI mismatch (e.g., between the CSI reconstructed at the NW side and the CSI estimated at the WTRU side). The WTRU may then determine a set of RIs and CQIs, e.g., one RI and CQI for each selected precoder method, and report the compressed full-channel CSI and the set RIs and CQIs to the network. In one optional embodiment, the report may include an indication of the one or more selected precoder methods used by the WTRU.

[0115] As used herein, the terms AE model, AI / ML model, ML model, and AI model may be used interchangeably to refer to models used for CSI compression. Furthermore, CSI mismatch may refer to a mismatch between a subset of CSI determined based on the input CSI of the WTRU-side model and a subset of CSI estimated or calculated based on the output CSI of the gNB-side model. CSI mismatch and input / output CSI mismatch may be used interchangeably herein. Mismatch may be referred to as the difference between the input of an autoencoder model and the output of the autoencoder model, where the encoder of the autoencoder (AE) may be deployed in the WTRU and the decoder may be deployed in the gNB. Mismatch and input / output mismatch, as well as NW (network) and gNB, may also be used interchangeably herein.

[0116] RI / CQI Configuration for AI / ML CSI Feedback. In various embodiments, the WTRU may be configured, determined, or indicated to report CSI feedback using a two-sided AI / ML model (e.g., an autoencoder) to compress the complete CSI feedback or a subset thereof. The CSI feedback may include, but is not limited to, a measured or predicted channel matrix, eigenvectors of the measured or predicted channel matrix, associated RI and / or CQI, associated L1 measurements (e.g., L1-RSRP, L1-SINR, LI, etc.), and associated PMI. One or more examples may include the following: (i) An AI / ML model for compression may be used for a subset of the CSI feedback, such as the channel matrix or a processed form of the channel matrix (e.g., eigenvectors of the channel matrix, a precoding matrix index associated with the channel matrix), and other portions of the CSI feedback may be reported without the compression performed by the AI / ML model. (ii) One or more AI / ML models may be used for compression and for the AI / ML model selected to use for CSI reporting, and may be determined based on the values ​​of a subset of CSI feedback (e.g., RI / CQI). For example, when the RI value is higher than a threshold (e.g., RI > threshold), a first AI / ML model may be used, and otherwise a second AI / ML model may be used. And (iii) a preprocessing scheme may determine the type of input data for the AI / ML model for compression, and may be determined based on the values ​​of a subset of CSI feedback (e.g., RI / CQI).The input data types may include, but are not limited to, a measured channel matrix, a predicted channel matrix, a first type of processed form of a channel matrix (e.g., eigenvectors), a second type of processed form of a channel matrix (e.g., a precoded channel matrix), a third type of processed form of a channel matrix (e.g., approximated to a precoding matrix), a fourth type of processed form of a channel matrix (e.g., a precoding matrix index), and related types.

[0117] In one embodiment, the WTRU may be provided with information for determining one or more of the CSI reporting quantities (e.g., PMI, CQI, RI, LI, etc.) when an AI / ML model is used to compress and / or predict the CSI to report to the gNB. This information may be provided by the gNB or may be predetermined based on a particular AI / ML model. The information for the WTRU may include:

[0118] Codebook information. For example, if the channel matrix is ​​used as an input for an AI / ML model (e.g., the WTRU-side model of a two-sided model), codebook information for determining the CQI / RI may be provided to the WTRU. As an example, the codebook information may include the codebook type (e.g., Type I, Type II, eigenvector to be reported), codebook configuration parameters (e.g., scaling factor, number of beams, codebook structure, etc.), and codebook subset constraint information.

[0119] Interference measurement resource (IMR) information. For example, the WTRU may be provided with interference measurement resources for CQI / RI determination.

[0120] Additional information for the WTRU may include a channel measurement resource (CMR), a highest rank, the number of eigenvectors to be reported, a subband size, an uplink resource to use for reporting, one or more AI / ML models, where the WTRU may determine the AI / ML model based on a determined CSI feedback size (or a determined feedback overhead), and / or one of more precoder computation methods, where the WTRU may report one or more RI / CQI and CSI feedback reports based on the precoder computation method.

[0121] Configuration for Handling CSI Feedback Mismatch. In one exemplary embodiment, the WTRU may be provided with information to estimate, calculate, derive, and / or determine a level of mismatch of CSI feedback (e.g., a subset of CSI feedback) at the WTRU side when a two-sided AI / ML model is used. The information for determining the level of CSI mismatch between the WTRU and the gNB may be provided by the network (e.g., via higher layer signaling or dynamic signaling) and may include one or more of the following:

[0122] -For a two-sided AI / ML model, the WTRU may already have the WTRU-side model (e.g., the first part of the two-sided AI / ML model), so that the WTRU may be provided with the gNB-side model (e.g., the second part of the two-sided AI / ML model) so that the WTRU may perform the decompression portion at the WTRU.

[0123] A threshold value for determining whether the WTRU needs to implement procedures to mitigate CSI mismatch.

[0124] A threshold for triggering WTRU behavior predetermined or configured by the network to mitigate CSI mismatch (e.g., switching / reselecting / activating / deactivating AI / ML models).

[0125] A subset of CSI feedback that should be monitored by the WTRU.

[0126] Uplink resources (e.g., PUCCH, PUSCH, Sounding Reference Signal (SRS)) for reporting supplemental information for mitigating CSI mismatch. In an example embodiment, the supplemental information may include a level or value of mismatch (e.g., a gap between the CQI / RI calculated based on the input channel matrix and the CQI / RI calculated based on the output channel matrix from a two-sided AI / ML model), an indication of whether a CSI mismatch mitigation procedure or scheme should be used, a report from the WTRU side of whether a CSI mismatch mitigation procedure or scheme is recommended, and / or an offset value to be used at the gNB side to mitigate CSI mismatch.

[0127] - A secondary AI / ML model to use when the CSI discrepancy is higher than a threshold.

[0128] - The AI / ML model to use based on the level of CSI discrepancy; and / or

[0129] A reference signal configuration (e.g., a precoded reference signal) for measuring, determining, deriving, or estimating the level of CSI mismatch. In one example, a precoded CSI-RS resource may be configured to measure CSI mismatch, and the WTRU may assume that the precoded CSI-RS is precoded using the reported CSI (or the most recent CSI report before the CSI reference timing).

[0130] In a modified embodiment, the WTRU may be configured to monitor CSI mismatch and / or perform CSI mismatch mitigation procedures, for example, when the WTRU reports a negative acknowledgement (NACK) N consecutive times, where N may be configured as a threshold value, and the WTRU observes a gap (e.g., SNR gap, MCS gap) greater than the threshold value between the scheduled MCS for the PDSCH and an estimated MCS based on channel measurements in the same slot (or adjacent time slots), and / or the WTRU is indicated to perform monitoring / mitigation procedures for CSI mismatch for a certain time period or time resource.

[0131] WTRU Procedures for Mismatch Mitigation. In some embodiments, the WTRU may select and / or apply mismatch (e.g., CSI mismatch) mitigation, for example, when the WTRU determines that a CSI mismatch event occurs or when the WTRU receives an indication from the NW to apply CSI mitigation. Example CSI mitigation methods may include requesting a switch to PUSCH if PUCCH was used for encoder output feedback, changing the compression ratio (e.g., lowering the compression ratio), switching to a different AI / ML encoder model, switching pre-processing, and / or reverting to a legacy CSI reporting method.

[0132] In some embodiments, the WTRU may switch to a data channel (e.g., PUSCH) for CSI feedback reporting, e.g., when one or more previous CSI reports used a control channel (e.g., PUCCH). The WTRU may decide to feed back the entire CSI or a subset of the CSI (e.g., a compressed channel matrix or compressed eigenvectors) over the data channel (e.g., PUSCH) in semi-persistent mode or aperiodically, depending on the configuration. The WTRU may send an indication to the NW requesting resources to report CSI over PUSCH, e.g., when semi-persistent or aperiodic reporting on PUSCH is not configured.

[0133] In another example, the WTRU may mitigate CSI mismatch by selecting a second compression rate for the AI / ML encoder, where the second compression rate is different (e.g., reduced compression) from the first compression rate used by the WTRU. The WTRU may select the second compression rate from a set of supported (e.g., configured) compression rates, possibly according to a predetermined rule or to meet a configured performance threshold. In one approach, the WTRU may select the highest compression supported by the AI / ML encoder if it meets a predetermined performance criterion (e.g., NMSE less than a threshold or SGCS greater than a threshold). In another embodiment, the WTRU may select the highest compression (e.g., less than the first compression rate) that meets the configured performance threshold and fits the configured reporting size.

[0134] When a WTRU is configured with a set of AI / ML encoder models, the WTRU may determine a second AI / ML encoder to use, e.g., to meet a configured performance threshold. In one example, the WTRU may select the lowest complexity AI / ML encoder to be paired with the NW-side AI / ML decoder and that meets a first (e.g., lowest) set of performance requirements, such as a first normalized mean square error (NMSE) threshold or a first square generalized cosine similarity (SGCS) threshold. In another approach, the WTRU may select the AI / ML encoder from the list of configured encoders to be paired with the NW-side AI / ML decoder and that has the best performance (e.g., NMSE or highest SGCS).

[0135] In a modified embodiment, the WTRU may mitigate CSI mismatch by switching pre-processing methods, including changing to a second pre-processing method or bypassing pre-processing. In one example, the WTRU may select a second pre-processing method and / or pre-processing parameters from a set of supported and / or configured pre-processing methods that results in a minimum AI / ML encoder model size and meets a configured performance threshold. For example, the WTRU may use a pre-processing method in the frequency domain and may determine to reduce the amount of averaging in the frequency domain to improve performance of the pre- / post-processing and AE pairs (e.g., reduce input / output mismatch). In another example, the WTRU may determine to bypass pre-processing when none of the supported / configured pre-processing methods and AE pairs meets a configured performance threshold.

[0136] A method for a WTRU to measure input / output CSI mismatch for a two-sided AI / ML model. In various embodiments, the input / output mismatch for a two-sided autoencoder (AE) model for an AI / ML encoder in a first node and an AI / ML decoder in a second node is evaluated. The first node may be a WTRU or a gNB, and the second node may be a gNB or a WTRU. In some embodiments, the measurement and detection are performed at the WTRU side and fed back to the gNB.

[0137] Configuration for mismatch detection in the WTRU. As previously discussed, the WTRU may be configured with multiple AIML encoder and decoder models for CSI feedback generation. The AIML encoder models are (E1,...,E N ), but the AIML decoder model can be written as (D1,...,D M), where N and M refer to the number of AIML encoder and decoder models available / accessible at the WTRU, respectively. Each AIML encoder model may be paired / compatible with one or more AIML decoder models, where model pairing / compatibility refers to encoder and decoder models that can work together (e.g., have the same input and output dimensions). For example, AIML encoder E1 has D1 and D2. M May be paired with / compatible with AIML encoder E N may be paired / compatible with the D1 and D2 decoder models. The WTRU may report its capabilities for the AIML decoder to the gNB.

[0138] In some embodiments, the WTRU may be configured to detect and report mismatches associated with a pre-configured encoder-decoder pair. In one example, the WTRU may be configured with a higher-layer parameter AECSIMIsmatch that takes on a binary value (0 or 1). For example, if AECSIMIsmatch is set to =1, the WTRU may calculate and / or report AE mismatch. The mismatch may be defined as how far the output of the configured AIML decoder is from the AIML encoder. The mismatch may represent the deviation of the decoder output from the encoder input, for example, due to compression loss. The AIML AE mismatch may be measured based on an indicated performance mismatch metric (e.g., normalized mean square error (NMSE) or generalized squared cosine similarity (GSCS)) between the input of the configured AIML encoder and the output of the configured AIML decoder model. The WTRU may be configured with a performance mismatch threshold parameter, e.g., AEMismatchThreshold, that takes on a set of scalar values ​​indicating a maximum acceptable mismatch threshold. For example, this threshold may represent an NMSE threshold that represents the maximum acceptable NMSE between the input and output of the configured encoder-decoder pair. If the calculated NMSE of the configured encoder-decoder pair is greater than a configured threshold, the WTRU may fall back to another strategy (eg, legacy) for deriving the CSI report.

[0139] In some embodiments, the WTRU may be configured with parameters (e.g., eigenvectors or full / raw channel state information) that indicate the mismatch region. For example, if the mismatch region is indicated as an eigenvector, the WTRU may derive the principal eigenvector associated with the input and output of the configured encoder-decoder pair and calculate the associated performance mismatch (e.g., NMSE) in the eigenvector domain.

[0140] The WTRU selects an encoder-decoder pair, e.g., (E1,D M) may be configured to select and report the configured AIML decoder model (e.g., D M ), where the selected encoder model together with the configured decoder satisfies a configured performance mismatch threshold.

[0141] In some embodiments, the WTRU may be configured to determine / detect a mismatch based on PDSCH performance. This may occur when the WTRU does not have access to any of the decoder models or when the WTRU does not have information about which decoder model is used on the gNB side. The WTRU may be configured with a block error rate (BLER) threshold or a certain number of consecutive NACK thresholds. For example, the WTRU may recommend switching AIML decoder and / or encoder models if the number of observed NACKs or BLER levels exceeds a configured threshold. In another embodiment, the WTRU may be configured with a threshold for the number of consecutive NACKs to mitigate AIML AE mismatch. For example, if the number of consecutive NACKs (N) exceeds a certain pre-configured threshold (e.g., N=4), the WTRU may recommend and report a different encoder model (e.g., with a larger feedback size) to reduce the mismatch error.

[0142] Configuration for Signaling Mismatch Measurement Region. The WTRU may be configured to measure and / or report a mismatch associated with a configured or selected two-sided AIML AE model. The WTRU may be configured with a higher layer parameter, e.g., AEMismatchDomain, that takes multiple values, each value indicating a region where the mismatch needs to be measured. For example, the mismatch may be measured in the eigenvector domain or in the raw CSI domain. In one option, the WTRU may be configured to measure and report the mismatch in the same input domain as the AIML AE input domain. In another option, the WTRU may be configured to measure the mismatch in a region different from that of the AIML AE input. For example, the gNB may configure the WTRU to measure and / or report the mismatch in the eigenvector domain for an AIML AE model trained in the raw CSI domain.

[0143] When configured to measure mismatch in the eigenvector domain, the WTRU may report the mismatch in a layer-specific mode or a layer-common mode. For example, when the WTRU is configured to report the mismatch in a layer-specific mode, the WTRU may measure and report the mismatch for each layer separately while in the layer-common mode, or the WTRU may measure the mismatch for all layers together, e.g., using an average value across all layers.

[0144] WTRU Measurements and Procedures for Mismatch Detection. The WTRU may be configured to measure and report mismatch of the two-sided AIML AE model. The mismatch may be measured based on a configured measure that depends on the inputs and outputs of the AE model. For example, if the WTRU is shown to measure the mismatch in the raw channel domain, the mismatch may be measured based on the function

[0145]

number

[0146] wherein

[0147]

number

[0148] and

[0149]

number

[0150] represent the input and output channels of the AIML encoder and decoder, respectively, and N r is the number of receive antenna ports in the WTRU, and N t is the number of transmit antenna ports at the gNB, and K is the number of configured subbands.

[0151]

number

[0152] can be one of the following:

[0153] The normalized mean square error (NMSE) may be calculated using Equation 1 below:

[0154]

number

[0155] where ||.|| F denotes the Frobenius norm of a 3D tensor. Alternatively, if configured or indicated, the WTRU may use Equation 2 to calculate the NMSE in the eigenvector domain as follows:

[0156]

number

[0157] where, for k=1,…K and l=1,…,L, w l [k] is the l-th right eigenvector of the k-th subband associated with H, while

[0158]

number

[0159] teeth

[0160]

number

[0161] where L is the maximum number of layers supported by the WTRU.

[0162] The weighted squared generalized cosine similarity (SGCS) and may be calculated using Equation 3 below:

[0163]

number

[0164] where, for k=1,…K and l=1,…,L, w l [k] is the l-th right eigenvector of the k-th subband associated with H, while

[0165]

number

[0166] teeth

[0167]

number

[0168] where L is the maximum number of layers supported by the WTRU. For l=1,...,L, the weight α l may be signaled by the gNB to the WTRU. l If is not signaled, the WTRU uses α l It may be assumed that σ should be set to 1 / L, i.e., unweighted SGCS. If the WTRU is configured to measure SGCS for layer eigenmodes,

[0169]

number

[0170] can be calculated using Equation 4 as follows:

[0171]

number

[0172] Here, the above function is calculated separately for each layer.

[0173] The WTRU may detect the occurrence of a mismatch by comparing the measured mismatch (e.g., based on SGCS or NMSE as previously described) with a corresponding pre-configured threshold. th When configured with

[0174]

number

[0175] ρ th and report discrepancies using Equation 5 as follows:

[0176]

number

[0177] In another option, the WTRU may determine whether the resulting discrepancy has a negative value, i.e., ρ mis < 0, it may be configured to report a mismatch. mis A positive value of ρ suggests that the constructed model is performing well. Alternatively, the WTRU may use ρ to indicate how well the inputs and outputs of the AIML AE model are correlated. mis As an alternative to measuring the mismatch based on the SGCS, the WTRU may be configured to measure and report the mismatch based on the NMSE using Equation 6 below.

[0178]

number

[0179] where ∈ mis , ∈ th and

[0180]

number

[0181] denotes the measured discrepancy, the configured NMSE threshold, and the NMSE measured by the WTRU. mis If it is <0, it may indicate that a mismatch occurs; otherwise, the AIML AE model is working properly.

[0182] In another embodiment, the WTRU may detect the occurrence of a mismatch by monitoring the number of consecutive NACKs or BLER levels. For example, the WTRU may compare the number N of consecutive NACKs with a preconfigured threshold N th Compared to N, th If more than 100,000, the discrepancy may be reported.

[0183] Mismatch Mitigation. In one exemplary embodiment, the WTRU may be configured with a method for mitigating mismatch associated with a two-sided AIML AE model. For example, if a mismatch is detected, the WTRU may recommend a different encoder and / or decoder, e.g., with a larger feedback size. In one option, the WTRU may recommend reducing the number of layers if the mismatch is dominated by some of the transmitted layers. For example, the WTRU may detect that the mismatch of the second layer is much larger than the first layer, so the WTRU may recommend rank-1 transmission. In another option, the WTRU may be configured to switch to a legacy method (e.g., CSI Type I or Type II) if the mismatch falls below a preconfigured minimum allowed mismatch level.

[0184] Feedback from the WTRU for mismatch detection. The WTRU may feed back an indication about the precoder mismatch to the gNB. In one method, the WTRU may be configured to report the mismatch, including the magnitude of the mismatch. The WTRU may also feedback the type of mitigation scheme applied at the WTRU side. As an example, the feedback about the mitigation scheme may indicate one or more mitigation schemes used, including requesting switching to PUSCH for CSI feedback, compression rate, AI / ML model, or information about preprocessing.

[0185] The WTRU may feedback a CSI report including compressed CSI, RI, CQI, and / or PMI. In one example, the CSI feedback size may change dynamically based on the mitigation scheme used in the WTRU. For example, if the WTRU decides to reduce the compression ratio by half, the bit width of the compressed CSI may increase accordingly, e.g., by a factor of two. In the case of a change in the bit width of the CSI report, the WTRU may request a CSI reporting resource configuration from the gNB.

[0186] In some embodiments, the decision on the mismatch mitigation scheme may be indicated by the gNB to the WTRU. If the mitigation scheme is indicated by the gNB, the WTRU may receive a CSI reporting configuration along with an indication of the mitigation scheme. The WTRU may be configured to transmit feedback via a channel indicated by the gNB, e.g., a PUSCH and / or a PUCCH.

[0187] Referring to FIG. 5, an example method 500 for a WTRU using a two-sided model for CSI feedback is shown. First, the WTRU may indicate its AI / ML model decoder capability to the network (505). Next, the WTRU is configured to measure / detect a mismatch between the reconstructed CSI at the NW side and the CSI estimated at the WTRU side (e.g., input / output CSI mismatch) (510). In one example, the WTRU includes a CSI mismatch measurement method, a CSI mismatch metric, and one or more thresholds for CSI mismatch detection. Examples of CSI mismatch metric 512, such as NMSE applied to a channel matrix or GSCS applied to an eigenvector associated with the channel matrix, were previously described.

[0188] The WTRU receives the CSI-RS from the gNB and estimates the CSI based on measurements of the CSI-RS. The WTRU measures or determines one or more values ​​of CSI mismatch based on a configured mismatch determination method and measurement scale (515). Examples of measuring or determining a CSI mismatch value 518 may include the WTRU using an AI / ML decoder to compare the input and output of a model at the gNB side to quantify the mismatch according to an indicated scale and / or monitoring transmission statistics to measure / observe the degree of CSI I / O mismatch. The WTRU may detect a CSI mismatch event when one or more measured or observed CSI mismatch values ​​are greater than or less than a first configured CSI mismatch decision threshold, depending on the method / scale / threshold used (520).

[0189] When the WTRU detects 520 a CSI mismatch event, the WTRU selects and applies 525 a CSI mismatch mitigation method depending on the input / output CSI mismatch measurement. As one example 528, if CSI feedback reporting is configured to use a data channel (PUSCH), the WTRU reduces the CSI feedback compression ratio when the input / output CSI mismatch measurement exceeds a second configured CSI mismatch detection threshold. Alternatively, if CSI feedback reporting is configured to use a control channel (PUCCH), the WTRU requests switching to data channel reporting (e.g., PUSCH) when the CSI mismatch measurement is lower than the second configured CSI mismatch detection threshold. Other examples 528 may include reducing the compression ratio of the CSI feedback, switching to a different AI / ML encoder model, switching or canceling pre-processing, and / or falling back to a legacy CSI reporting framework.

[0190] Finally, the WTRU reports the CSI feedback and CSI mismatch information to the gNB 530. In one embodiment, the CSI mismatch information includes input / output CSI mismatch measurements, an indication that a CSI mismatch event has occurred, and / or an employed or preferred CSI mismatch mitigation scheme.

[0191] In an embodiment for a WTRU procedure for NW-side input / output CSI mismatch detection and reporting for a two-sided AI / ML model, the input / output mismatch is determined for a two-sided autoencoder (AE) model including an AI / ML encoder at a first node and an AI / ML decoder at a second node, where the first node may be a WTRU or a gNB, and the second node may be a gNB or a WTRU. Measurement and detection may be performed on the gNB side using test vectors transmitted from the WTRU to the gNB.

[0192] WTRU configuration for mismatch detection at a gNB. In one exemplary embodiment, the WTRU may receive a configuration for a network-side mismatch detection procedure. The exemplary configuration may be signaled in an RRC message, for example, in an RRC setup and / or RRC reconfiguration message. Alternatively, such a configuration may be predefined, for example, as a default radio configuration.

[0193] In one exemplary embodiment, the WTRU configuration may include a test vector configuration. The WTRU may be configured to apply the test vector, or a portion thereof, as an input to an AI / ML model associated with CSI compression. The WTRU may be configured to transmit an output of the AI / ML model corresponding to the test vector input to the gNB. The transmission of the output of the AI / ML model corresponding to the input test vector may be considered WTRU feedback for mismatch detection at the gNB.

[0194] According to some embodiments, test vectors may be configured as standalone inputs to the AI / ML model. For example, the WTRU may apply the test vectors, or portions thereof, as inputs to the model. For example, the inputs to the AI / ML model may not include any channel matrix information. In one example, the WTRU may be pre-configured with a set of test vectors. In another example, the WTRU may be configured with rules for generating the test vectors. For example, the test vectors may be pseudo-random sequences. In various embodiments, the size / dimension of the test vectors may be equal to the input size / dimension of the AI / ML model. The WTRU may be configured with multiple test vectors, or in another example, the WTRU may be configured with a base test vector and multiple cyclic shifts of the base test vector. When multiple test vectors are configured, the WTRU may select one test vector based on one or more rules. For example, the WTRU may choose a test vector based on a function of frame and / or subframe and / or slot number. In another example, the WTRU may choose a test vector based on a CSI reporting configuration. The WTRU may be configured with a pseudo-random pattern to choose a test vector from multiple configured / generated test vectors. In another example, the WTRU may randomly choose a test vector based on the WTRU implementation.

[0195] In one embodiment, the test vectors may be configured as partial inputs to the AI / ML model. For example, the WTRU may apply as inputs to the model a portion of the input being the test vectors and the remaining portion being based on channel information (e.g., a channel matrix, an eigenvector, or any pre-processed version thereof). Similar to standalone test vectors, the WTRU may be pre-configured with a set of test vectors. In another embodiment, the WTRU may be configured with rules for generating test vectors. For example, the test vectors may be pseudo-random sequences. In one example, the size / dimension of the test vectors may be smaller than the input size / dimension of the AI / ML model. The WTRU may be configured with multiple test vectors, or may be configured with a base test vector and multiple cyclic shifts of the base test vector. When multiple test vectors are configured, the WTRU may select one test vector based on one or more rules. For example, the WTRU may choose a test vector based on a function of the frame and / or subframe and / or slot number. In another example, the WTRU may choose a test vector based on a CSI reporting configuration. In other embodiments, the WTRU may be configured with a pseudo-random pattern to select a test vector from a plurality of configured / generated test vectors, or the WTRU may select a test vector randomly based on the WTRU implementation.

[0196] In some approaches, the WTRU may be configured with a multiplexing rule between the test vectors and the channel information. For example, the WTRU may be configured to multiplex the test vectors and the channel information in a comb pattern. For example, given test vectors [t1, t2... tk, tk+1... tn] and channel information [c1, c2... cn], the WTRU may perform multiplexing such that the resulting input vector is [t1, t2... tk, c1, c2... cn, tk+1, tk+2... tn]. For example, given an input vector [1...N], the WTRU may be configured to multiplex test vectors in even positions and channel information in odd positions, or vice versa. For example, the WTRU may be configured to multiplex the test vectors according to a preconfigured pattern. For example, the preconfigured pattern may be generated by a pseudo-random generator. In another example, the preconfigured pattern may depend on the frame and / or subframe and / or slot number. In some examples, the preconfigured pattern may be configured by the gNB. In another example, the multiplexing pattern may depend on the CSI reporting configuration.

[0197] Test vector type selection based on CSI reporting instance. In some embodiments, the WTRU may be configured with both standalone test vectors and partial test vectors. The WTRU may be configured to determine the type of test vector to apply based on the CSI reporting instance. For example, if a test vector transmission collides / coincides with a CSI reporting instance, the WTRU may use a partial test vector. For example, if a test vector transmission does not collides / coincide with a CSI reporting instance, the WTRU may use a standalone test vector.

[0198] In one embodiment, the test vectors may be defined before preprocessing. For example, the WTRU may be configured to apply the same type of preprocessing to the test vectors and the channel information. In another approach, the test vectors may be defined after preprocessing. For example, the WTRU may be configured to apply preprocessing for the channel information but skip preprocessing for the test vectors.

[0199] Feedback from the WTRU for mismatch detection at the gNB may use one or more triggers for test vector transmission. Embodiments may be applicable to standalone test vector and / or partial test vector transmission. According to some embodiments, the WTRU may be configured to periodically transmit test vectors based on a preconfigured period. The period of the test vector transmission, if configured, may be an integer multiple of the periodic CSI report. For example, the WTRU may be configured to transmit a test vector every N transmissions of a CSI report, where the value of N may be preconfigured.

[0200] In one embodiment, the WTRU may be configured to transmit a test vector when a preconfigured condition is met. As an example, the WTRU may be configured to transmit a test vector when the number of NACKs (possibly consecutive NACKs) within a preconfigured period exceeds a threshold. In another example, the WTRU may be configured to transmit a test vector when the difference in CQI / PMI / RI between consecutive CSI reports exceeds a threshold. In another example, the WTRU may be configured to transmit a test vector when the difference between the reported CQI and the MCS allocated by the gNB exceeds a preconfigured threshold. In yet another example, the WTRU may be configured to transmit a test vector(s) when it determines that a change in channel conditions (e.g., channel coherence time, channel coherence bandwidth) within a preconfigured period exceeds a certain threshold.

[0201] In some embodiments, the WTRU may be configured with dedicated UL resources for test vector transmission. For example, the UL resources may be PUCCH resources and / or PUSCH resources. In an example, the WTRU may be configured to transmit the test vectors on resources configured for CSI reporting. The WTRU may be configured to transmit additional information along with the test vector transmission. This additional information may depend on the type of UL resource allocated for the test vector transmission. For example, if the WTRU is allocated PUSCH resources for test vector transmission, the WTRU may transmit only the test vector transmission. In some embodiments, if the WTRU is allocated PUCCH resources for test vector transmission, the WTRU may transmit both the input to the encoder and the output of the encoder associated with the test vector. Various combinations are possible.

[0202] According to some embodiments, a WTRU procedure for CSI mismatch mitigation based on a gNB indication may include the WTRU receiving an indication from the gNB of a mismatch between a precoder calculated by the WTRU and a precoder determined by the gNB. The indication from the gNB may be in response to WTRU feedback of a test vector. In one embodiment, the indication from the gNB may be in response to WTRU feedback of a mismatch detection. In other embodiments, the indication from the gNB may be based on a mismatch detection at the gNB. The WTRU may be configured to perform one or more mitigation actions upon receiving the mismatch indication from the gNB. In one embodiment, the mismatch indication from the gNB may further configure the WTRU to perform a particular mitigation procedure. Some examples of mitigation procedures may include (i) requesting a switch to PUSCH if PUCCH was used for encoder output feedback, (ii) changing the compression ratio (e.g., lowering the compression ratio), (iii) switching to a different AI / ML encoder model, (iv) switching or undoing pre-processing, and / or (v) reverting to legacy processing.

[0203] 6, a method 600 for a WTRU configured to use a two-sided model for CSI feedback and support NW-side measurement and detection of mismatch between NW-side reconstructed CSI and WTRU-side estimated CSI (input / output CSI mismatch) is shown. In one exemplary embodiment, the WTRU receives 605 configuration information including, for example, a test vector type (standalone, partial input, input to pre-processing, or input to AI / ML encoder), a set of test vectors or pre-configured patterns, test vector selection criteria, metrics to monitor, and / or one or more CSI mismatch mitigation methods.

[0204] The WTRU is triggered by at least one of a time (e.g., based on a preconfigured period and offset) and an event (e.g., based on a monitored metric exceeding a configured threshold) to transmit test vectors to the NW for measuring CSI mismatch (610). The WTRU selects one or more test vector types and / or test vectors based on the test vector selection criteria and the monitored metric according to its configuration. Example test vector types 612 may include standalone test vectors that cover the entire input of the AI / ML model and / or partial test vectors that cover indicated portions of the input based on multiplexing rules.

[0205] The WTRU may transmit the selected test vectors to the NW (610) and report the compressed CSI to the gNB (620). If and / or when the WTRU receives a CSI mismatch indication from the NW (625), the WTRU selects and / or applies a configured CSI mismatch mitigation method (630). As previously mentioned, in some examples 632, the WTRU may reduce the CSI compression rate, switch to a different AI / ML encoder model, switch or disable pre-processing, and / or the WTRU requests switching to PUSCH (e.g., if PUCCH was used) for CSI feedback reporting. In one example, the WTRU then reports the selected mismatch mitigation method to the NW (635) and transmits CSI feedback to the NW based on the selected CSI mismatch mitigation method.

[0206] Referring to FIG. 7, a method 700 for updating RI / CQI for eigenvector-based AI / ML CSI feedback is shown and may include updating (or fine-tuning) one or both of RI / CQI to mitigate autoencoder input / output mismatch for eigenvector-based CSI compression, under the assumption that the WTRU does not have a gNB decoder.

[0207] In method 700, a base station determines and transmits configuration information and reference signals (RS), e.g., CSI-RS, to a WTRU (702) to enable autoencoder mismatch determination. The WTRU receives the configuration and RS and, based thereon, estimates the channel (H) (705) and calculates a precoder (w) and initial RI / CQI, which are transmitted to the base station to determine (706) a configuration and updated rank for UL reporting of W by the WTRU, which are transmitted back to the WTRU. Using the received UL reporting configuration and updated rank, the WTRU transmits (709) a compressed precoder W to the base station and reconstructs the precoder on the NW side (710). Based on the reconstructed precoder, the base station may schedule DL resources (712) and transmit a precoded RS and, optionally, DL data to the WTRU (714). Based on the received precoded RS, the WTRU may determine an effective precoder gain 715, update the RI / CQI, and send the updated RI / CQI to the base station. Based on the received updated RI / CQI, the base station may schedule DL resources with the updated RI / CQI 718 and send DL data to the WTRU using the scheduled DL resources 720.

[0208] In the WTRU procedure for determining an updated RI / CQI, initial RI and CQI determination 705 is performed. The WTRU may be configured with a precoding method to be used for downlink transmission. The configuration for the precoding method may indicate the use of methods such as eigenvector decomposition, singular value decomposition, zero-forcing, maximum ratio transmission, etc. The WTRU may calculate the rank and channel quality based on the configured precoding method. For example, the calculation of the rank and channel quality may be based on first finding the rank with the highest total SINR and then matching the SINR with channel quality, for example, using a lookup table.

[0209] For a given precoding method, the WTRU may calculate the corresponding precoding gain 715. As an example, the WTRU may calculate the HH H Eigenvector decomposition method may be used to calculate the precoder w, where w=λw, H represents the channel matrix, and λ represents the original precoding gain of the corresponding layer.

[0210] In reporting AI / ML CSI feedback, the WTRU may request uplink resources from the gNB for CSI feedback. In the case of eigenvector-compressed AI / ML CSI feedback, the size of the CSI feedback that the WTRU feeds back to the gNB, e.g., the number of eigenvectors, may depend on the rank. The WTRU may feed back the RI and CQI and request an uplink reporting configuration from the gNB before transmitting the CSI feedback. The WTRU may receive the uplink reporting configuration for CSI feedback. In another option, the uplink reporting configuration may be implicit to the WTRU and the gNB. The WTRU may also receive an indication of the rank from the gNB, which may be different from what is reported by the WTRU. Based on the uplink reporting configuration, the WTRU transmits AI / ML CSI feedback. In the case of eigenvector-based CSI feedback, the WTRU feeds back compressed eigenvectors for each of the layers.

[0211] In some embodiments, the WTRU may be configured to receive a precoded reference signal, which may be used to measure the effective precoding gain and update the RI / CQI. According to various embodiments, the precoded reference signal may be based on the following example options:

[0212] (i) Precoded CSI-RS. Conventional CSI-RS used to measure the channel is not precoded. Precoded CSI-RS may be precoded using a precoder determined / reconstructed by the gNB. In one example, the WTRU may receive the precoded CSI-RS and use them to measure the effective precoding gain and update the RI / CQI.

[0213] (ii) Demodulation Reference Signal (DMRS). Traditionally, DMRS is a precoded reference signal used for channel estimation. In one embodiment, the WTRU may reuse the DMRS to extract the effective precoding gain and update the RI / CQI. And / or

[0214] (iii) Precoding Monitoring Reference Signals. In one method, the WTRU may receive a set of precoded reference signals designed to measure the effective precoding gain at the WTRU, which may be used to update the RI / CQI. This type of reference signal may be additional to existing reference signals. The precoding monitoring reference signals may occupy time-frequency resources that include a period as indicated by the gNB.

[0215] As shown in FIG. 7, to update 715 the RI / CQI, the WTRU may measure the precoding gain using the precoded reference signal.

[0216]

number

[0217] The precoder w computed by the WTRU may be reconstructed by the gNB.

[0218]

number

[0219] The received reference signal r is given by

[0220]

number

[0221] is precoded using

[0222]

number

[0223] where s denotes the sequence of the reference signal and (·) H denotes the Hermitian operator, H denotes the channel matrix, and n denotes additive noise. The WTRU may calculate the effective precoding gain based on the received precoded RS r using one of the following example options:

[0224] Option a: Given by Equation 7 (autocorrelation-based method):

[0225]

number

[0226] Option b: Given by Equation 8 (cross-correlation based method):

[0227]

number

[0228] In option a, the received reference signal is used to calculate the effective precoding gain, which means that the effective precoding gain and the noise power P nThis option may be used when the RI / CQI update process in the WTRU does not have access to the combiner v.

[0229] In option b, the received reference signal is correlated with a combiner v known to the WTRU. This option may be used when the RI / CQI updating process in the WTRU has access to combiner v.

[0230] The WTRU may be configured with a threshold value for determining for precoder mismatch. In one embodiment, the WTRU may compare the effective precoding gain to the original precoding gain to determine for precoder mismatch based on the example options given below. The selection of the effective precoding gain calculation method may be indicated by the gNB or selected by the WTRU. In one option for comparing λ and λ′, the WTRU may compare the original precoding gain λ to an effective precoding gain λ′ calculated using a cross-correlation based method. λ and

[0231]

number

[0232] In another option, the WTRU compares the original precoding gain λ with the effective precoding gain λ calculated using an autocorrelation-based method.

[0233]

number

[0234] can be compared to

[0235] The WTRU may be configured to calculate and report a new RI / CQI corresponding to the effective precoding gain. In one method, the WTRU may calculate the rank to find the highest total SINR using the effective precoding gain. The WTRU may then calculate the channel quality by matching the SINR with the channel quality, for example, using a lookup table.

[0236] As shown in method 700 of FIG. 7, after receiving the conventional CSI-RS (705), the WTRU may feedback the RI and CQI as part of a CSI reporting configuration request. The WTRU may then receive a CSI reporting configuration and rank from the gNB (709). The WTRU may report CSI feedback on the configured uplink resources. The WTRU may then receive a downlink scheduling configuration along with a precoded reference signal, calculate a new RI / CQI based on the precoded reference signal (715), and report the updated (or second) RI / CQI.

[0237] In various embodiments for feedback from the WTRU for RI / CQI and CSI, the WTRU may indicate a request for a CSI feedback configuration to the gNB. The requested CSI feedback configuration may depend on the rank determined by the WTRU. The WTRU may be configured with resources, e.g., layers, on which to report the precoder method used for one or more precoders corresponding to each rank. In some embodiments, the CSI feedback report may include at least one of the precoder (e.g., the WTRU may report eigenvectors for the channel matrix, or compressed eigenvectors, or the output of the precoder matrix encoder), PMI, RI, and / or CQI. In one embodiment, the WTRU may report secondary CSI feedback to report a new RI CQI value.

[0238] 8, a method 800 for a WTRU using a two-sided model for CSI feedback is configured to report RI / CQI if it determines that an input / output CSI mismatch event has occurred 805. An example WTRU configuration includes parameters for implementing eigenvector (EV)-based CSI compression, a threshold for determining an input / output CSI mismatch event (e.g., a precoding gain threshold), and a reporting configuration for compressed CSI feedback.

[0239] The WTRU receives the CSI-RS, determines the CSI (including a first RI and CQI) (810), calculates the original precoding gain, and performs EV-based CSI compression. The WTRU may request (815) and receive (820) a UL reporting configuration based on the initial rank to report the first RI and CQI and the compressed CSI (e.g., a first precoder or precoder matrix, or an indication thereof, associated with the first determined RI and CQI). The WTRU receives (825) a precoded reference signal (RS) using a second precoder (e.g., the second precoder is determined at the gNB and may be different from the first precoder). Examples of the precoded reference signal 826 have been previously described and may be a precoded CSI-RS, a DMRS, or a precoded monitoring reference signal. The WTRU then measures 830 an effective precoding gain based on the received RS precoded with the second precoder, where the effective precoding gain is the gain of the precoded channel. Examples 832 for determining the effective precoding gain have been previously described and may include autocorrelation or cross-correlation methods. The WTRU may determine a second / updated RI and CQI when the difference between the measured effective precoding gain and the original precoding gain exceeds a configured precoding gain threshold. The WTRU reports 835 the second / updated RI and CQI.

[0240] In an additional embodiment, a method for selecting a precoder method for AI / ML CSI compression based on full channels using a two-sided AI / ML model is disclosed. In an exemplary embodiment, a WTRU procedure for supporting multiple precoder methods is included. In one example, a WTRU may be configured to use one or more of a set of precoder methods to determine one or more precoder matrices. The precoders may include SVD, ZF, and MRT.

[0241] The WTRU may determine values ​​of a rank indicator (RI) and a CQI that depend on the precoder method. For example, the WTRU may perform measurements on a set of RSs and determine a channel matrix. The WTRU may use a first precoder method to determine a first precoder matrix applicable to the measured channel matrix. For this first precoder matrix, the WTRU may determine an associated first RI and a first CQI. The WTRU may use a second precoder method to determine a second precoder matrix applicable to the measured channel matrix. For this second precoder matrix, the WTRU may determine an associated second RI and a second CQI.

[0242] In an example embodiment for selecting a precoder method, the WTRU may determine the number of precoder methods and the identity of the precoder methods for which to provide CSI feedback. The WTRU may select one or more precoder methods for which to obtain PMI and associated RI / CQI and report CSI feedback. In various embodiments, the selection may be based on at least one of the following:

[0243] Configuration by the gNB. For example, the WTRU may receive an indication or configuration for a set of one or more precoder methods for which to report CSI feedback. In another example, the WTRU may be configured to report CSI feedback for up to x WTRU-selected precoder methods (x may be configurable). The configuration may be received in RRC. The configuration may be received dynamically (e.g., in DCI), for example, in case of triggering aperiodic or semi-persistent CSI reporting.

[0244] The precoder method that yields the highest CQI or RI. For example, the WTRU may rank the different precoder methods in order of CQI value and choose to report CSI feedback for the top x precoder methods.

[0245] - Mismatches associated with precoder methods. For example, the WTRU may select one or more precoder methods based on previously determined mismatches associated with the precoder methods. For example, the WTRU may select x precoder methods with the lowest mismatch. In another example, the WTRU may select y precoder methods whose mismatch is less than a (possibly configurable) threshold.

[0246] Compression rate of the coded CSI feedback. For example, the WTRU may select a precoder method based on the achievable compression rate of the CSI feedback encoder.

[0247] - Transmission performance. For example, the WTRU may select or decide to change the precoder method based on previous transmission performance (e.g., ACK / NACK rate).

[0248] Measurements. For example, the WTRU may be configured with measurements and triggers on measurements to enable selection of a precoder method. Some precoder methods may provide better performance (e.g., mismatch range, transmission performance, feedback payload) in certain channel conditions.

[0249] - Transmission requirements. For example, the WTRU may select a precoder method based on the required latency or reliability of the associated transmission.

[0250] Feedback payload. For example, the WTRU may select a precoder method based on the CSI feedback payload. In one example, the WTRU may determine the number of precoder methods for which it should provide feedback depending on the CSI feedback payload.

[0251] Reception of precoded data using a precoded RS or precoder method. For example, the WTRU may select a first precoder method for providing CSI feedback depending on the precoded RS or precoder method used to generate the precoded data. The WTRU may be triggered to report a CSI feedback report depending on determining that the precoded RS does not result in the same value (e.g., RI or CQI) of a previously fed back CSI report.

[0252] According to some embodiments, the WTRU may be triggered to select or reselect one or more precoder methods based on at least one of: (i) a determination that a discrepancy relative to previously reported CSI feedback is greater than a (possibly configurable) threshold; (ii) performance of one or more associated transmissions (e.g., the WTRU may be triggered to reselect one or more precoder methods based on the HARQ-ACK performance of the associated transmissions, where the associated transmissions may be transmissions for which a previous precoder method was used to precode the transmission); (iii) an indication from the gNB (e.g., the indication may include a discrepancy value); (iv) measurements (e.g., the WTRU may determine that one or more previously selected precoders are no longer applicable to new channel conditions); and / or (v) measurements performed on a precoded RS (e.g., the WTRU may determine an RI or CQI based on measurements performed on the precoded RS). The WTRU may determine that the RI or CQI differs from the previously reported RI or CQI by more than an offset value. In this case, the WTRU may report updated RI and / or CQI values. The WTRU may also be prompted to select a new precoder method, which may be the one whose associated PMI or RI or CQI best matches that used for the precoded RS.

[0253] In some embodiments, the WTRU may be configured with a default or fallback precoder method. In one example, the WTRU may receive an indication (e.g., from the gNB) of when to use the default precoder method. In another example, the WTRU may decide to use the default precoder method when the mismatch exceeds a threshold. The WTRU may always report CSI feedback for the default precoder method.

[0254] Referring to FIG. 9, an example signaling diagram is shown for a method 900 for updating RI / CQI in AI / ML CSI feedback based on a channel matrix. The base station determines and sends configuration information and CSI-RS to the WTRU. After receiving the conventional CSI-RS, the WTRU may estimate CSI for channel H (905) and determine a set of initial RI / CQIs according to configuration by the NW. The WTRU may calculate RI / CQIs based on the configuration of the precoder. The NW reconstructs the precoder based on how the WTRU was configured. In method 900, the WTRU reconstructs the precoder according to the WTRU configuration, schedules DL resources, and transmits the determined set of RI / CQIs and the compressed channel matrix (H) as part of CSI feedback for the base station to transmit a precoded RS for the reconstructed precoder, optionally along with DL data, for example. The WTRU may then receive the downlink scheduling configuration and the precoded reference signal, calculate an effective precoder gain, update the RI / CQI, and select a precoder with an optimal effective precoder gain (915). The WTRU may calculate a new RI / CQI based on the precoded reference signal and report the updated RI / CQI and preferably an indication of the selected precoder. The base station determines DL scheduling with the updated RI / CQI and transmits DL data to the WTRU based on the DL scheduling with the updated RI / CQI.

[0255] Example embodiments also relate to feedback from a WTRU for multiple precoder methods. The WTRU may indicate the selected one or more precoder methods to the gNB and may be configured with resources on which to report the precoder methods used for one or more CSI feedback reports. In some examples, the WTRU may include one or more precoder method identities used for CSI feedback in a CSI feedback report instance. An example CSI feedback report may include at least one of: (i) identification of one or more precoder methods; (ii) the number of precoder methods used; (iii) a channel matrix (e.g., the WTRU may report a measured channel matrix, or a compressed channel matrix, or the output of a channel matrix encoder); (iv) PMI; (v) RI; and / or (vi) CQI.

[0256] In one method, the WTRU may report at least one channel matrix and / or PMI and / or RI and / or CQI for each selected precoder method. In another method, the WTRU may select more than one precoder method but report the channel matrix and / or PMI and / or RI and / or CQI for a subset of the selected precoder methods (e.g., for a single precoder method). In another method, the WTRU may determine at least one of the PMI, RI, or CQI based on one or more selected precoder methods. For example, the WTRU may determine the RI or CQI that leads to the best performance (e.g., lowest mismatch) for any of the selected precoder methods. The WTRU may request new feedback reporting resources when selecting one or more precoder methods. For example, the WTRU may request larger feedback reporting resources if the WTRU selects more precoder methods than were used for the previous feedback reporting instance.

[0257] 10 , an example method 1000 for a WTRU using a two-sided model for CSI feedback is shown, where the WTRU is configured to select one or more precoder methods for determining and reporting compressed CSI or RI / CQI. In one example, the WTRU receives 1005 configuration information including a set of precoder methods (e.g., SVD, ZF) for determining a precoder, one or more precoder method selection thresholds, and one or more reporting configurations for compressed CSI feedback and RI / CQI.

[0258] In method 1000, the WTRU receives 1010 the CSI-RS and performs CSI compression of the entire channel matrix. The WTRU determines one or more precoder methods depending on the reporting configuration, the measured channel conditions, and / or one or more precoder method selection thresholds. Example thresholds include the WTRU selecting (i) the precoder method that results in the highest CQI or RI, (ii) a precoder method based on measurements, and / or (iii) a precoder method based on a determined input / output CSI mismatch (between the CSI reconstructed at the NW side and the CSI estimated at the WTRU side).

[0259] The WTRU then determines 1015 a set of RIs and CQIs, e.g., one RI and CQI, for each selected precoder method and selects 1020 a precoder method. Examples of selecting a precoder method 1022 have been previously described and may be based on, for example, statistical performance, an indication by the gNB, measurements on precoded RSs precoded with multiple precoders, transmission requirements, and / or feedback payload. The WTRU reports 1025 a compressed full-channel CSI and a set RI and CQI based on the selected precoding method. In one example, the report may include an indication of one or more selected precoder methods. Finally, the WTRU receives 1030 the precoded RSs and calculates / feeds back new RIs / CQIs when a precoding mismatch is detected.

[0260] 11 , an example method 1100 for a WTRU to measure input / output CSI mismatch for a two-sided AI / ML model may generally include the WTRU receiving 1105 CSI mismatch configuration information to configure the WTRU with one or more CSI mismatch measurement methods, one or more CSI mismatch measurement measures, and one or more thresholds for CSI mismatch detection. The WTRU receives 1110 a CSI reference signal and estimates CSI. The WTRU may determine 1115 an input / output mismatch based on the measurement and the configured mismatch measure. If 1120 the determined mismatch exceeds a configured mismatch threshold, a CSI mismatch event is detected 1125, and the WTRU performs 1130 CSI mismatch mitigation. In some embodiments, the CSI mismatch mitigation is performed dependent on the input / output mismatch measurement.

[0261] In embodiments using the WTRU-side CQI adjustment method, when the WTRU does not have an accurate model of CSI reconstruction or does not receive precoded RS, the WTRU needs to estimate and report the CQI as part of the CSI feedback. The following embodiments address the case where the WTRU calculates the CQI based on real channel measurements using CSI-RS, applies the adjustment parameter Δ, and reports the CQI accordingly.

[0262] The WTRU may be configured for a CQI adjustment parameter Δ, which may be a fixed value. The parameter Δ may be transparent to the NW or may be indicated to the WTRU by the NW.

[0263] The WTRU may be configured to use a method to calculate Δ. The method for calculating Δ may be transparent to the NW or may be indicated to the WTRU by the NW.

[0264] The WTRU may be configured with a method for measuring the mismatch, and may be further configured to report the mismatch feedback as part of an incremental MCS based link adaptation procedure.

[0265] The WTRU may be configured to report CQI only, CQI-Δ only, or both CQI and Δ separately as part of the CSI report.

[0266] The WTRU may be configured with a method for estimating CQI.

[0267] In this embodiment, there may be various procedures for link level adaptation including CQI adjustment and fixed CQI adjustment, link adaptation using CQI adjustment and / or CQI estimation methods as further detailed below.

[0268] In an embodiment using a fixed CQI adjustment, the WTRU may be configured to report a CQI based on a fixed adjustment parameter Δ. First, the WTRU calculates the CQI based on actual channel measurements using the CSI-RS and reports CQI-Δ as a CQI value as part of the CSI feedback. In another alternative, the WTRU may be configured with a method for calculating the fixed adjustment parameter Δ. First, the WTRU calculates the CQI based on actual channel measurements. The adjustment parameter is determined based on a configured method. As an example, the WTRU may be configured to calculate Δ as a percentage of the calculated CQI. As another example, the adjustment parameter may be calculated based on a function of parameters such as the channel matrix, the WTRU velocity, the AE model, etc. After calculating the adjustment parameter Δ based on the configured method, the WTRU reports CQI-Δ as a CQI value as part of the CSI feedback.

[0269] 12, a method 1200 for determining a mismatch using link adaptation with CQI adjustment is shown. A WTRU may be configured with a link adaptation procedure based on CQI adjustment, a method for calculating an adjustment parameter Δ, and separately reporting the CQI and Δ as part of CSI feedback (1205). First, the WTRU calculates the CQI based on actual channel measurements. Then, the WTRU calculates the adjustment parameter Δ based on the shown method. The WTRU reports the calculated CQI and Δ separately as part of CSI feedback (1210).

[0270] The WTRU uses the index MCS calculated based on the CQI-Δ. iThe WTRU receives 1215 a data transmission with an MCS scheme denoted by min(MCS max ,MCS i+1 ) and the data transmission may be received using MCS parameters 1230 calculated based on the MCS max denotes the maximum MCS value that can be obtained with CQI calculated using real channel measurements, and MCS i+1 is MCS i Refers to an MCS that is one level higher than the standard MCS.

[0271] In case of ACK and NOK message feedback (i.e., a mismatch event is detected), the WTRU may select the MCS despite the ACK message to prevent mismatch during the next data transmission. i MCS parameter 1213MCS, which is one level lower than i-1 Additionally, to prevent a ping-pong effect between the ACK / OK and ACK / NOK cases, when an ACK / NOK case occurs, the WTRU may receive the data transmission using MCS i-1 In the case of NACK feedback, a retransmission occurs and the MCS calculated based on CQI-Δ is used. i Since the MCS is reset every time the OK and NOK messages become outdated.

[0272] In an embodiment using a CQI estimation method, the WTRU may be configured to estimate the CQI using an analytical model of the CQI reconstructed at the NW side. For example, the analytical model CQI of the CQI calculated at the NW side NW WTRU, CQI NW , SINR, and CQI NW Based on the observations, the CQI NW The estimation function for may be defined as using the following equation:

[0273]

number

[0274] In another option, the WTRU may be configured to use a trained model of CQI reconstruction on the NW side. For example, the model may be configured such that the inputs to the model are the channel matrix and SINR, and the output is the CQI. NW As another example, the inputs to the model are the eigenvectors and SINR, and the output is the CQI NW After training the model, the WTRU can obtain the CQI at the NW side, i.e.,

[0275]

number

[0276] A model can be used to estimate

[0277] In another option, the WTRU may be configured to use a modified CSI reconstruction part (i.e., a decoder of an autoencoder) that differs from the actual CSI reconstruction part at the NW to calculate the CQI. As an example, the trained model may include a lower number of trainable parameters to save storage and computation at the WTRU side. The WTRU may use the modified CSI reconstruction model to estimate the reconstructed CSI at the NW. The WTRU may then calculate the CQI. NW , i.e.

[0278]

number

[0279] The estimated CSI may be used in the NW to estimate .

[0280] According to the configuration regarding the CQI estimation method, the WTRU:

[0281]

number

[0282] The adjustment parameter Δ is calculated so that

[0283] In embodiments using feedback from the WTRU for CQI adjustment, the WTRU may feed back a different CQI field to the NW as part of the CSI feedback based on one of CQI-Δ in a single CQI field or CQI and Δ in separate fields. Δ represents the level difference between the CQI calculated at the WTRU and the estimated CQI at the NW side. Δ may have a maximum bit width of the size dedicated to reporting the conventional CQI field.

[0284] Referring to FIG. 13 , an example method 1300 for a WTRU to mitigate CSI mismatch detection for a two-sided AI / ML model may generally include the WTRU receiving 1305 configuration information to support NW-side measurement and detection of mismatch between NW-side reconstructed CSI and WTRU-side estimated CSI (input / output CSI mismatch). The example configuration includes a test vector type (standalone, partial input, input to preprocessing, or input to an AI / ML encoder), a set of test vectors or preconfigured patterns, test vector selection criteria, a metric to monitor, and one or more CSI mismatch mitigation methods. The WTRU is triggered 1310 to transmit test vectors to the NW by at least one of time (e.g., based on a preconfigured period and offset) or an event (e.g., based on a monitored metric exceeding a configured threshold). The WTRU selects one or more test vector types and / or test vectors based on the test vector selection criteria and the monitored metric, and transmits 1310 the selected test vectors to the NW. If the WTRU receives a CSI mismatch indication from the NW (1315), the WTRU selects and / or applies a configured CSI mismatch mitigation method (1320). One example mitigation method may include the WTRU reducing the CSI compression rate. In other examples, the WTRU switches to a different AI / ML encoder model, or the WTRU switches or disables pre-processing, or the WTRU requests switching to PUSCH for CSI feedback reporting (e.g., if PUCCH was used). The WTRU may transmit CSI feedback to the NW based on the selected CSI mismatch mitigation method.

[0285] 14, an example method 1400 for a WTRU to update an RI / CQI for eigenvector-based AI / ML CSI feedback using a two-sided AI / ML model may generally include the WTRU receiving 1405 configuration information to report an RI / CQI if it determines that an input / output CSI mismatch event has occurred. The example configuration includes parameters for performing eigenvector (EV)-based CSI compression, one or more thresholds (e.g., precoding gain thresholds) for determining an input / output CSI mismatch event, and one or more reporting configurations for compressed CSI feedback. The WTRU receives 1410 a CSI-RS, determines CSI (including a first RI and CQI), calculates an original precoding gain, and performs EV-based CSI compression. The WTRU reports 1415 the first RI and CQI and reports the compressed CSI (e.g., a first precoder or precoder matrix, or an indication thereof, associated with the first determined RI and CQI). The WTRU receives 1420 a reference signal (RS) precoded with a second precoder (e.g., the second precoder is determined at the gNB and may be different from the first precoder) and measures an effective precoding gain based on the received RS precoded with the second precoder, where the effective precoding gain is the gain of the precoded channel. The WTRU then determines 1430 a second RI and CQI if the difference between the measured effective precoding gain and the original precoding gain exceeds a configured precoding gain threshold, and the WTRU reports the second RI and CQI.

[0286] 15 , an example method 1500 for a WTRU to select a precoder method for full-channel-based AI / ML CSI compression using a two-sided AI / ML model may generally include receiving 1505 configuration information for the WTRU to select one or more precoder methods for determining and reporting compressed CSI or RI / CQI. The example configuration includes a set of precoder methods (e.g., SVD, ZF) for determining a precoder, one or more precoder method selection thresholds, and one or more reporting configurations for compressed CSI feedback and RI / CQI. The WTRU receives 1510 a CSI-RS and performs CSI compression of the full channel matrix. The WTRU determines 1515 one or more precoder methods depending on the reporting configuration, measured channel conditions, and / or one or more precoder method selection thresholds. In various examples, the WTRU selects the precoder method that results in the highest CQI or RI based on measurements or based on a determined input / output CSI mismatch (between the CSI reconstructed at the NW side and the CSI estimated at the WTRU side). The WTRU then determines 1520 a set of RIs and CQIs, one RI and CQI for each selected precoder method, and reports 1525 the compressed full-channel CSI and the determined set of RIs and CQIs. Optionally, the report may include an indication of one or more selected precoder methods.

[0287] Although features and elements have been described above in particular combinations, those skilled in the art will understand that each feature or element may be used alone or in any combination with the other features and elements. In addition, the methods described herein may be implemented in a computer program, software, or firmware embodied in a computer-readable medium for execution by a computer or processor. Examples of computer-readable media include electronic signals (transmitted over wired or wireless connections) 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, cache memory, semiconductor memory devices, magnetic media such as internal hard disks and removable disks, magneto-optical media, and optical media such as CD-ROM disks and digital versatile disks (DVDs). A processor may be used in conjunction with software to implement a radio frequency transceiver for use in a WTRU, UE, terminal, base station, RNC, or any host computer.

Claims

1. 1. A method for a wireless transmit / receive unit (WTRU), comprising: receiving, from a base station, configuration information for detecting input / output (I / O) channel state information (CSI) mismatch of a two-sided artificial intelligence and machine learning (AI / ML) model, the configuration information including a CSI mismatch measurement method, a CSI mismatch metric, and one or more thresholds for CSI mismatch detection; receiving one or more CSI reference signals (CSI-RSs); and estimating CSI based on the received CSI-RSs; determining an I / O CSI mismatch measurement based on at least one of the CSI mismatch metric, the estimated CIA, or the configured CSI mismatch measurement method; determining a CSI mismatch event when the determined I / O CSI mismatch measurement exceeds a first CSI mismatch detection threshold of the configured one or more thresholds for CSI mismatch detection; selecting a CSI mismatch mitigation method as a function of the I / O CSI mismatch measurement; reporting CSI feedback to the base station about the received CSI-RS and CSI mismatch information, the CSI mismatch information including at least one of the I / O CSI mismatch measurement, an indication of the determined CSI mismatch event, or the selected CSI mismatch mitigation method; A method comprising:

2. the selected CSI mismatch mitigation method is Identifying whether the reporting of the CSI feedback is transmitted using a Physical Uplink Shared Channel (PUSCH) or a Physical Uplink Control Channel (PUCCH); (i) if reporting the CSI feedback using the PUSCH, reducing a CSI feedback compression rate when the I / O CSI mismatch measurement value exceeds a second CSI mismatch detection threshold among the one or more configured thresholds for CSI mismatch detection, or (ii) if reporting the CSI feedback using the PUCCH, requesting switching to transmitting the CSI feedback to the PUSCH when the I / O CSI mismatch measurement value is equal to or less than a second CSI mismatch detection threshold among the one or more configured thresholds for CSI mismatch detection; The method of claim 1 , comprising:

3. 2. The method of claim 1 , wherein the selected CSI mismatch mitigation method comprises selecting and reporting a pair of encoders and decoders that yields I / O CSI measurements that are lower than the first CSI mismatch detection threshold.

4. The method of claim 1 , wherein the CSI mismatch metric comprises a normalized mean squared error (NMSE) or a weighted squared generalized cosine similarity (SGCS) of a channel with the base station.

5. The method of claim 1 , wherein the determined I / O CSI mismatch measurement is based on transmission statistics of previous CSI feedback.

6. The method of claim 5 , wherein the transmission statistics include a number of consecutive negative acknowledgements (NACKs).

7. 2. The method of claim 1, wherein reporting the CSI feedback for the received CSI-RS includes compressed CSI, a rank indicator (RI), a channel quality index (CQI), and a recoding matrix index (PMI).

8. 1. A wireless transmit / receive unit (WTRU), comprising: a transceiver and a processor communicatively coupled to the transceiver, the transceiver and processor comprising: receiving, from a base station, configuration information for detecting input / output (I / O) channel state information (CSI) mismatch of a two-sided artificial intelligence and machine learning (AI / ML) model, the configuration information including a CSI mismatch measurement method, a CSI mismatch metric, and one or more thresholds for CSI mismatch detection; receiving one or more CSI reference signals (CSI-RSs); and estimating CSI based on the received CSI-RSs; determining an I / O CSI mismatch measurement based on at least one of the CSI mismatch metric, the estimated CIA, or the configured CSI mismatch measurement method; determining a CSI mismatch event when the determined I / O CSI mismatch measurement exceeds a first CSI mismatch detection threshold of the configured one or more thresholds for CSI mismatch detection; selecting a CSI mismatch mitigation method as a function of the I / O CSI mismatch measurement; reporting CSI feedback to the base station about the received CSI-RS and CSI mismatch information, the CSI mismatch information including at least one of the I / O CSI mismatch measurement value, an indication of the determined CSI mismatch event, or the selected CSI mismatch mitigation method; The WTRU is configured as follows.

9. The selected CSI mismatch mitigation method comprises the transceiver and processor: Identifying whether the reporting of the CSI feedback is transmitted using a Physical Uplink Shared Channel (PUSCH) or a Physical Uplink Control Channel (PUCCH); (i) if the reporting of the CSI feedback uses the PUSCH, reducing a CSI feedback compression rate when the I / O CSI mismatch measurement value exceeds a second CSI mismatch detection threshold among the one or more configured thresholds for CSI mismatch detection, or (ii) if the reporting of the CSI feedback uses the PUCCH, requesting switching to transmitting the CSI feedback to the PUSCH when the I / O CSI mismatch measurement value is equal to or less than a second CSI mismatch detection threshold among the one or more configured thresholds for CSI mismatch detection. The WTRU of claim 8 , further configured to:

10. 10. The WTRU of claim 8, wherein the selected CSI mismatch mitigation method comprises the transceiver and processor being further configured to select and report an encoder and decoder pair that results in an I / O CSI measurement that is lower than the first CSI mismatch detection threshold.

11. The WTRU of claim 8 , wherein the CSI mismatch metric comprises a normalized mean squared error (NMSE) or a weighted squared generalized cosine similarity (SGCS) of a channel with the base station.

12. The WTRU of claim 8 , wherein the determined I / O CSI mismatch measurement is based on transmission statistics of previous CSI feedback.

13. The WTRU of claim 12 , wherein the transmission statistics include a number of consecutive negative acknowledgements (NACKs).

14. 10. The WTRU of claim 8, wherein the reporting of the CSI feedback for the received CSI-RS includes a compressed CSI, a rank indicator (RI), a channel quality index (CQI), and a recoding matrix index (PMI).

15. A base station, a transceiver and a processor communicatively coupled to the transceiver, the transceiver and processor comprising: transmitting, to a wireless transmit / receive unit (WTRU), configuration information for detecting input / output (I / O) channel state information (CSI) mismatch of a two-sided artificial intelligence and machine learning (AI / ML) model, the configuration information including a CSI mismatch measurement method, a CSI mismatch metric, and one or more thresholds for CSI mismatch detection; the WTRU transmitting one or more CSI reference signals (CSI-RS) for estimating CSI; receiving a rank indicator (RI) and a channel quality index (CQI) from the WTRU based on the one or more SCI-RCs; receiving the transmitted CSI-RS and CSI mismatch information from the WTRU, the received CSI mismatch information including at least one of an I / O CSI mismatch measurement, an indication of a determined CSI mismatch event, or a selected CSI mismatch mitigation method; The base station is configured as follows.

16. The base station of claim 15 , wherein the selected CSI mismatch mitigation method comprises a recommended encoder and decoder pair that results in an I / O CSI measurement that is lower than the first CSI mismatch detection threshold.

17. 16. The base station of claim 15, wherein the CSI mismatch metric in the transmitted configuration information comprises a normalized mean squared error (NMSE) or a weighted squared generalized cosine similarity (SGCS) of the channel.

18. 16. The base station of claim 15, wherein the receiving CSI feedback for the transmitted CSI-RS includes compressed CSI, a rank indicator (RI), a channel quality index (CQI), and a recoding matrix index (PMI).

Images