Measurement and reporting of metrics for performance monitoring

Abstract

Systems, methods, and tools for measuring and reporting metrics for performance monitoring are disclosed herein. A device may receive configuration information indicating an input data type, a set of reference vectors, and a distance metric type. The apparatus may receive a channel state information (CSI) reference signal (CSI-RS). The device may determine a measured CSI based on a measurement value associated with the CSI-RS. The device may generate a compressed CSI based on the measured CSI and the input data type. The device may calculate a distance metric associated with the measured CSI and a reference vector of the set of reference vectors based on the distance metric type. The device may send a report to a network node. The report may indicate a compressed CSI and a distance metric associated with the measured CSI and a reference vector in a set of reference vectors.

Description

[0001] Cross-references to related applications

[0002] This application claims the benefit of U.S. Provisional Patent Application No. 63 / 456,758, filed April 3, 2023, the contents of which are incorporated herein by reference. Background Technology

[0003] Mobile communication using wireless communication continues to evolve. The fifth generation can be called 5G. Previous generations (traditional) mobile communication can be, for example, fourth generation (4G) Long Term Evolution (LTE). Summary of the Invention

[0004] The systems, methods, devices, and tools described in this article relate to the measurement and reporting of metrics for performance monitoring.

[0005] A device (e.g., a Wireless Transmitter / Receiver Unit (WTRU)) can be configured to receive configuration information indicating the input data type, a set of reference vectors, and a distance metric. The device can receive a Channel State Information (CSI) Reference Signal (CSI-RS). The device can determine a measured CSI based on measurements associated with the CSI-RS. The device can generate a compressed CSI based on the measured CSI and the input data type. The device can calculate the distance associated with the measured CSI and reference vectors in the set of reference vectors based on the distance metric. The device can send a report to a network node. This report can indicate the compressed CSI and distance associated with the reference vectors in the set of measured CSI and reference vectors.

[0006] The device can calculate the distance associated with a reference vector in the set of measured CSIs and reference vectors by calculating the corresponding distance between each reference vector in the set of measured CSIs and reference vectors.

[0007] The distance metric can be the normalized mean square error (NMSE). The device can calculate the distance associated with a reference vector in the set of measured CSIs and reference vectors by calculating the NMSE of the reference vector in the set of measured CSIs and reference vectors.

[0008] The distance metric can be cosine similarity. The device can calculate the distance associated with a reference vector in the set of measured CSI and reference vectors by computing the cosine similarity between the measured CSI and the reference vectors in the set of reference vectors.

[0009] Computing the distance associated with the measured CSI and a reference vector of the set of reference vectors can involve computing respective distances between the measured CSI and each reference vector of the set of reference vectors, and wherein the report further indicates a minimum distance of the respective distances or a maximum distance of the respective distances.

[0010] The distance metric can be a function that maps an input tensor to a scalar value. The distance associated with the measured CSI and a reference vector of the set of reference vectors can be a scalar value.

[0011] The reference vector is associated with a first domain, and the distance metric can be associated with a second domain. The device can transform the reference vector from the first domain to the second domain. The input data type can be a full channel matrix or an eigenvector. BRIEF DESCRIPTION OF DRAWINGS

[0012] Figure 1A FIG. 1 is a system diagram illustrating an example communications system in which one or more disclosed embodiments can be implemented.

[0013] Figure 1B FIG. 2 is a system diagram illustrating an example wireless transmit / receive unit (WTRU) that can be used within the communications system illustrated in FIG. 1 according to an embodiment. Figure 1A

[0014] Figure 1C FIG. 3 is a system diagram illustrating an example radio access network (RAN) and an example core network (CN) that can be used within the communications system illustrated in FIG. 1 according to an embodiment. Figure 1A

[0015] Figure 1D FIG. 4 is a system diagram illustrating a further example RAN and a further example CN that can be used within the communications system illustrated in FIG. 1 according to an embodiment. Figure 1A

[0016] Figure 2 FIG. 5 illustrates an example of a bilateral artificial intelligence / machine learning (AI / ML) based channel state information (CSI) compression framework.

[0017] Figure 3 FIG. 6 illustrates an example set of reference vectors, measured CSI, and a distance metric.

[0018] Figure 4 FIG. 7 illustrates an example of determining a set of reference vectors.

[0019] Figure 5 FIG. 8 illustrates an example of determining distances, subband groups, and quantization levels. DETAILED DESCRIPTION

[0020] ​​​Figure 1A is a diagram illustrating an example communications system 100, in which one or more disclosed embodiments can be implemented. The communications system 100 can be a multiple access system that provides content, such as voice, data, video, messaging, broadcast, etc., to multiple wireless users. The communications system 100 can enable multiple wireless users to access such content through the sharing of system resources, including wireless bandwidth. For example, the communications systems 100 can employ one or more channel access methods, such as code division multiple access (CDMA), time division multiple access (TDMA), frequency division multiple access (FDMA), orthogonal FDMA (OFDMA), single-carrier FDMA (SC-FDMA), zero-tail unique-word DFT-Spread OFDM (ZT UW DTS-s OFDM), unique word OFDM (UW-OFDM), resource block-filtered OFDM, filter bank multicarrier (FBMC), and the like.

[0021] As shown in Figure 1A The communications system 100 can include wireless transmit / receive units (WTRUs) 102a, 102b, 102c, 102d, a RAN 104 / 113, a CN 106 / 115, a public switched telephone network (PSTN) 108, the Internet 110, and other networks 112, as shown in FIG. 1. However, it is to be understood that the disclosed embodiments contemplate any number of WTRUs, base stations, networks, and / or network elements. Each of the WTRUs 102a, 102b, 102c, 102d can 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 can be referred to as a “station” and / or a “STA”) can be configured to transmit and / or receive wireless signals, and can include a user equipment (UE), a mobile station, a fixed or mobile subscriber unit, a subscription-based unit, a pager, a cellular telephone, a personal digital assistant (PDA), a smartphone, a laptop, a netbook, a personal computer, a wireless sensor, a hotspot or Mi-Fi device, an Internet of Things (loT) device, a watch or other wearable, a head-mounted display (HMD), a vehicle, a drone, a medical device and applications (e.g., remote surgery), an industrial device and applications (e.g., a robot or other wireless devices operating in an industrial and / or an automated processing chain context), a consumer electronics device, a device operating on a business and / or industrial wireless network, and the like. Any of the WTRUs 102a, 102b, 102c, and 102d can be interchangeably referred to as a UE.

[0022] The communications system 100 can also include a base station 114a and / or a base station 114b. Each of the base stations 114a, 114b can be any type of device configured to wirelessly interface to at least one of the WTRUs 102a, 102b, 102c, 102d to facilitate access to one or more communication networks, such as the CN 106 / 115, the Internet 110, and / or the other networks 112. By way of example, the base stations 114a, 114b can be a base transceiver station (BTS), a Node-B, an eNode B, a Home Node B, a Home eNode B, a gNB, a NR NodeB, a site controller, an access point (AP), a wireless router, and the like. While the base stations 114a, 114b are each depicted as a single element, it will be appreciated that the base stations 114a, 114b can include any number of interconnected base stations and / or network elements.

[0023] The base station 114a can be part of the RAN 104 / 113, which can 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 the base station 114b can be configured to transmit and / or receive wireless signals on one or more carrier frequencies (which can be referred to as a cell (not shown)). These frequencies can be in the licensed spectrum, the unlicensed spectrum, or a combination of licensed and unlicensed spectrum. A cell can provide service to a particular geographic area and this can be a relatively fixed geographic area or can change from time to time. Further, the cell can be further divided into cell sectors each with a coverage area. For example, the cell associated with the base station 114a can be divided into three sectors. Thus, in one embodiment, the base station 114a can include three transceivers, one for each sector of the cell. In an embodiment, the base station 114a can employ multiple-input multiple-output (MIMO) techniques and can utilize multiple transceivers for each sector of the cell. For example, beamforming can be used to transmit and / or receive signals in a desired spatial direction.

[0024] The base stations 114a, 114b can communicate with one or more of the WTRUs 102a, 102b, 102c, 102d over the air interface 116, which can be any suitable wireless communication link (e.g., radio frequency (RF), microwave, centimeter wave, micrometer wave, infrared (IR), ultraviolet (UV), visible light, etc.). The air interface 116 can be established using any suitable radio access technology (RAT).

[0025] More specifically, as noted above, the communications system 100 can be a multiple access system and can employ one or more channel access schemes, such as CDMA, TDMA, FDMA, OFDMA, SC-FDMA, and the like. For example, the base station 114a and WTRUs 102a, 102b, 102c in the RAN 104 / 113 can implement a radio technology such as Universal Mobile Telecommunications System (UMTS) Terrestrial Radio Access (UTRA), which can establish the air interface 115 / 116 / 117 using wideband CDMA (WCDMA). WCDMA can include communication protocols such as High-Speed Packet Access (HSPA) and / or Evolved HSPA (HSPA+). HSPA can include High-Speed Downlink (DL) Packet Access (HSDPA) and / or High-Speed UL Packet Access (HSUPA).

[0026] In an embodiment, the base station 114a and the WTRUs 102a, 102b, 102c can implement a radio technology such as Evolved UMTS Terrestrial Radio Access (E-UTRA), which can establish the air interface 116 using Long Term Evolution (LTE) and / or LTE-Advanced (LTE-A) and / or LTE-Advanced Pro (LTE-A Pro).

[0027] In an embodiment, the base station 114a and the WTRUs 102a, 102b, 102c can implement a radio technology such as NR Radio Access, which can establish the air interface 116 using New Radio (NR).

[0028] In an embodiment, the base station 114a and the WTRUs 102a, 102b, 102c can implement multiple radio access technologies. For example, the base station 114a and WTRUs 102a, 102b, 102c can implement LTE wireless access together with NR wireless access, for instance using dual connectivity (DC) principles. Thus, the air interface utilized by WTRUs 102a, 102b, 102c can be characterized by multiple types of radio access technologies and / or transmissions sent to / from multiple types of base stations (e.g., a eNB and a gNB).

[0029] In other embodiments, the base station 114a and the WTRUs 102a, 102b, 102c can implement radio technologies such as IEEE 802.11 (i.e., Wireless Fidelity (WiFi), IEEE 802.16 (i.e., Worldwide Interoperability for Microwave Access (WiMAX)), CDMA2000, CDMA2000 IX, 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), and the like.

[0030] Figure 1A The base station 114b in the FIG. 10 example can be a wireless router, Home Node B, Home eNode B, or access point, for example, and can utilize any suitable RAT for facilitating wireless connectivity access points for local area network(s) (LAN) 112, such as mesh networks, Wi-Fi networks, wireless TV networks, and / or the like. The base station 114b and the WTRUs 102c, 102d in the FIG. 10 example can 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 can 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 can utilize a cellular-based RAT (e.g., WCDMA, CDMA2000, GSM, LTE, LTE-A, LTE-A Pro, NR, etc.) to establish a picocell or femtocell. As Figure 1A As shown in the FIG. 10 example, the base station 114b can have a direct connection to the Internet 110. Thus, the base station 114b can not be required to access the Internet 110 via the CN 106 / 115.

[0031] The RAN 104 / 113 can be in communication with the CN 106 / 115, which can be any type of network configured to provide voice, data, applications, and / or voice over internet protocol (VoIP) services to one or more of the WTRUs 102a, 102b, 102c, 102d. The data can have varying quality of service (QoS) requirements, such as differing throughput requirements, latency requirements, error tolerance requirements, reliability requirements, data throughput requirements, mobility requirements, and the like. The CN 106 / 115 can provide call control, billing services, mobile location-based services, pre-paid calling, Internet connectivity, video distribution, etc., and / or perform high-level security functions, such as user authentication. Although not shown in the FIG. 10, a Figure 1AAs shown, but to be understood, RAN104 / 113 and / or CN 106 / 115 can communicate directly or indirectly with other RANs that use the same RAT as or a different RAT than RAN 104 / 113. For example, in addition to being connected to RAN 104 / 113, which can utilize NR radio technology, CN 106 / 115 can also communicate with another RAN (not shown) that uses GSM, UMTS, CDMA 2000, WiMAX, E-UTRA, or WiFi radio technology.

[0032] CN 106 / 115 can also be used as a gateway for WTRU 102a, 102b, 102c, 102d to access PSTN 108, the Internet 110, and / or other networks 112. PSTN 108 may include a circuit-switched telephone network providing Common Old-Style Telephone Service (POTS). The Internet 110 may include a global system of interconnected computer networks and devices using common communication protocols such as Transmission Control Protocol (TCP), User Datagram Protocol (UDP), and / or Internet Protocol (IP) from the TCP / IP Internet Protocol suite. Network 112 may include wired and / or wireless communication networks owned and / or operated by other service providers. For example, network 112 may include another CN connected to one or more RANs, which may use the same RAT as RAN 104 / 113 or a different RAT.

[0033] Some or all of the WTRUs 102a, 102b, 102c, and 102d in communication system 100 may include multi-mode capabilities (e.g., WTRUs 102a, 102b, 102c, and 102d may include multiple transceivers for communicating with different wireless networks via different wireless links). For example, Figure 1A The WTRU 102c shown can be configured to communicate with base station 114a, which can employ cellular-based radio technology, and with base station 114b, which can employ IEEE 802 radio technology.

[0034] Figure 1B This is a system diagram illustrating example WTRU 102. (Example:) Figure 1B As shown, among other things, WTRU 102 may also include 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 supply 134, a Global Positioning System (GPS) chipset 136, and / or other peripheral devices 138. It will be understood that, while remaining consistent with the embodiments, WTRU 102 may include any sub-combination of the foregoing elements.

[0035] The processor 118 can be a general purpose processor, a special purpose processor, a conventional processor, a digital signal processor (DSP), a plurality of microprocessors, one or more microprocessors in association with a DSP core, a controller, a microcontroller, Application Specific Integrated Circuits (ASICs), Field Programmable Gate Array (FPGAs) circuits, any other type of integrated circuit (IC), a state machine, and the like. The processor 118 can 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 can be coupled to the transceiver 120, which can be coupled to the transmit / receive element 122. While Figure 1B The processor 118 and the transceiver 120 are depicted as separate components, it is to be understood that the processor 118 and the transceiver 120 can be integrated together in an electronic package or chip.

[0036] The transmit / receive element 122 can be configured to transmit signals to, or receive signals from, a base station (e.g., the base station 114a) over the air interface 116. For example, in one embodiment, the transmit / receive element 122 can be an antenna configured to transmit and / or receive RF signals. In an embodiment, the transmit / receive element 122 can 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 can be configured to transmit and / or receive both RF and light signals. It will be appreciated that the transmit / receive element 122 can be configured to transmit and / or receive any combination of wireless signals.

[0037] Although the transmit / receive element 122 is depicted in the WTRU 102 Figure 1B In one embodiment, the WTRU 102 can include two or more transmit / receive elements 122 (e.g., multiple antennas) to enable MIMO technology. Thus, the WTRU 102 can be configured to transmit wireless signals using MIMO technology.

[0038] The transceiver 120 can be configured to modulate information to be transmitted by the transmit / receive element 122 and to demodulate information received by the transmit / receive element 122. As indicated above, the WTRU 102 can be a multi-mode device. Thus, the transceiver 120 can include multiple transceivers for enabling the WTRU 102 to communicate via multiple RATs, such as NR and IEEE 802.11, for example.

[0039] The processor 118 of the WTRU 102 can be coupled to, and can receive user input data from, the speaker / microphone 124, the keypad 126, and / or the display / touchpad 128 (e.g., a liquid crystal display (LCD) display unit or organic light-emitting diode (OLED) display unit). The processor 118 can also output user data to the speaker / microphone 124, the keypad 126, and / or the display / touchpad 128. In addition, the processor 118 can access information from, and store data in, any type of suitable memory, such as the non-removable memory 130 and / or the removable memory 132. The non-removable memory 130 can include random-access memory (RAM), read-only memory (ROM), a hard disk, or any other type of memory storage device. The removable memory 132 can include a subscriber identity module (SIM) card, a memory stick, a secure digital (SD) memory card, and the like. In other embodiments, the processor 118 can access information from, and store data in, memory that is not physically located on the WTRU 102, such as on a server or a home computer (not shown).

[0040] The processor 118 can receive power from the power source 134 and can be configured to distribute and / or control the power to the other components in the WTRU 102. The power source 134 can be any suitable device for powering the WTRU 102. For example, the power source 134 can include one or more dry cell batteries (e.g., nickel-cadmium (NiCd), nickel-zinc (NiZn), nickel metal hydride (NiMH), lithium-ion (Li-ion), etc.), solar cells, fuel cells, and the like.

[0041] The processor 118 can also be coupled to the GPS chipset 136, which can be configured to provide location information (e.g., longitude and latitude) regarding the current location of the WTRU 102. In addition to, or in lieu of, the information from the GPS chipset 136, the WTRU 102 can receive location information over the air interface 116 from a base station (e.g., base stations 114a, 114b) and / or determine its location based on timers from the arrival times of

[0042] The processor 118 may be further coupled to other peripheral devices 138, which may include one or more software and / or hardware modules providing additional features, functions, and / or wired or wireless connectivity. For example, peripheral devices 138 may include accelerometers, electronic compasses, satellite transceivers, digital cameras (for photos and / or video), Universal Serial Bus (USB) ports, vibration devices, television transceivers, hands-free headsets, etc. Modules, FM radio units, digital music players, media players, video game player modules, internet browsers, virtual reality and / or augmented reality (VR / AR) devices, activity trackers, and the like. Peripheral device 138 may include one or more sensors, which may be one or more of the following: gyroscope, accelerometer, Hall effect sensor, magnetometer, orientation sensor, proximity sensor, temperature sensor, time sensor; geolocation sensor; altimeter, light sensor, touch sensor, magnetometer, barometer, gesture sensor, biometric sensor, and / or humidity sensor.

[0043] WTRU 102 may include a full-duplex radio for which the transmission and reception of some or all signals (e.g., associated with specific subframes for both UL (e.g., for transmission) and downlink (e.g., for reception)) may be concurrent and / or simultaneous. The full-duplex radio may include an interference management unit for reducing and / or substantially eliminating self-interference through signal processing via hardware (e.g., a choke) or via a processor (e.g., a separate processor (not shown) or via processor 118). In an embodiment, WTRU 102 may include a half-duplex radio for which the transmission and reception of some or all signals (e.g., associated with specific subframes for UL (e.g., for transmission) or downlink (e.g., for reception)) may be concurrent and / or simultaneous.

[0044] Figure 1C The diagram illustrates a system diagram of RAN 104 and CN 106 according to an embodiment. As noted above, RAN 104 can communicate with WTRUs 102a, 102b, and 102c via air interface 116 using E-UTRA radio technology. RAN 104 can also communicate with CN 106.

[0045] The RAN 104 can include eNode-Bs 160a, 160b, 160c, though it will be appreciated that the RAN 104 can include any number of eNode-Bs while remaining consistent with an embodiment. The eNode-Bs 160a, 160b, 160c can each include one or more transceivers for communicating with the WTRUs 102a, 102b, 102c over the air interface 116. In one embodiment, the eNode-Bs 160a, 160b, 160c can implement MIMO technology. Thus, the eNode-B 160a, for example, can use multiple antennas to transmit wireless signals to, and / or receive wireless signals from, the WTRU 102a.

[0046] Each of the eNode-Bs 160a, 160b, 160c can be associated with a particular cell (not shown) and can be configured to handle radio resource management decisions, handover decisions, scheduling of users in the UL and / or DL, and the like. As shown, the eNode-Bs 160a, 160b, 160c can communicate with one another over an X2 interface. Figure 1C

[0047] Figure 1C The CN 106 can include a mobility management entity (MME) 162, a serving gateway (SGW) 164, and a packet data network (PDN) gateway (or PGW) 166, as shown. While each of the foregoing elements are depicted as part of the CN 106, it will be appreciated that any of these elements can be owned and / or operated by an entity other than the CN operator.

[0048] The MME 162 can be connected to each of the eNode-Bs 162a, 162b, 162c in the RAN 104 via an S1 interface and can serve as a control node. For example, the MME 162 can be responsible for authenticating users of the WTRUs 102a, 102b, 102c, bearer activation / deactivation, selecting a particular serving gateway during an initial attach of the WTRUs 102a, 102b, 102c, and the like. The MME 162 can 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.

[0049] ​The SGW 164 can be connected to each of the eNode Bs 160a, 160b, 160c in the RAN 104 via the S1 interface. The SGW 164 can generally route and forward user data packets to / from the WTRUs 102a, 102b, 102c. The SGW 164 can perform other functions, such as anchoring user planes during inter-eNode B handovers, triggering paging when DL data is available for the WTRUs 102a, 102b, 102c, managing and storing contexts of the WTRUs 102a, 102b, 102c, and the like.

[0050] The SGW 164 can be connected to the PGW 166, which can 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.

[0051] The CN 106 can also serve as a gateway for the WTRUs 102a, 102b, 102c to access the PSTN 108, the Internet 110, and / or the other networks 112. The PSTN 108 can include circuit-switched telephone networks that provide

[0052] Although WTRUs are described in Figures 1A-1D representative embodiments as wireless terminals, it is contemplated that in certain representative embodiments such terminals can (e.g., temporarily or permanently) use wired communication interfaces with the communication network.

[0053] In representative embodiments, the other network 112 can be a WLAN.

[0054] A WLAN in Infrastructure Basic Service Set (BSS) mode can have an Access Point (AP) for the BSS and one or more stations (STAs) associated with the AP. The AP can have an access or an interface to a Distribution System (DS) or another type of wired / wireless network that carries traffic in to and / or from the BSS. Traffic to STAs that originates from outside the BSS can arrive through the AP and can be delivered to the STAs. Traffic originating from STAs to destinations outside the BSS can be sent to the AP to be delivered to respective destinations. Traffic between STAs within the BSS can be sent through the AP, for example, where the source STA can send traffic to the AP and the AP can deliver the traffic to the destination STA. The traffic between STAs within a BSS can be considered and / or referred to as peer-to-peer traffic. Peer-to-peer traffic can be sent between (e.g., directly between) the source and destination STAs with a direct link setup (DLS). In certain representative embodiments, the DLS can use an 802.11e DLS or an 802.11z tunneled DLS (TDLS). A WLAN using an Independent BSS (IBSS) mode can not have an AP, and / or the STAs (e.g., all of the STAs) within or using the IBSS can communicate directly with each other. The IBSS mode of communication can sometimes be referred to herein as “ad-hoc” mode of communication.

[0055] When using an 802.11 ac infrastructure mode of operation or similar modes of operation, an AP can transmit beacons on a fixed channel, such as a primary channel. The primary channel can be a fixed width (e.g., 20 MHz wide bandwidth) or a width that is dynamically set via signaling. The primary channel can be the operating channel of the BSS and can be used by STAs to establish a connection with the AP. In certain representative embodiments, a carrier sense multiple access with collision avoidance (CSMA / CA) with collision avoidance can be implemented, for example, in 802.11 systems. For CSMA / CA, a STA, including the AP, (e.g., each STA) can sense the primary channel. If the primary channel is sensed / detected as busy and / or determined to be busy by a particular STA, the particular STA can backoff. Only one STA can transmit in the given BSS at any given time.

[0056] High Throughput (HT) STAs can use 40 MHz wide channels for communication, for example, via a combination of the primary 20 MHz channel with an adjacent or nonadjacent 20 MHz channel to form a 40 MHz wide channel.

[0057] Very High Throughput (VHT) STAs can support 20MHz, 40 MHz, 80 MHz, and / or 160 MHz wide channels. 40 MHz and / or 80 MHz channels can be formed by combining contiguous 20 MHz channels. A 160 MHz channel can be formed by combining 8 contiguous 20 MHz channels, or by combining two non-contiguous 80 MHz channels, which can be referred to as an 80+80 configuration. For the 80+80 configuration, after channel coding, the data can be parsed by a segment parser that can separate the data into two streams. Inverse Fast Fourier Transform (IFFT) processing and time domain processing can be done separately on each stream. The streams can be mapped on to the two 80 MHz channels, and the data can be transmitted by the transmitting STA. At the receiver of the receiving STA, the above described operations for the 80+80 configuration can be reversed, and the combined data can be sent to the Medium Access Control (MAC).

[0058] 802.11af and 802.11ah support sub-1 GHz modes of operation. The channel operating bandwidth and carriers are reduced in 802.11af and 802.11ah relative to those used in 802.11η and 802.11ac. 802.11af supports 5 MHz, 10 MHz, and 20 MHz bandwidths in TV White Space (TVWS) spectrum, and 802.11ah supports 1 MHz, 2 MHz, 4 MHz, 8 MHz, and 16 MHz bandwidths using non-TVWS spectrum. According to representative embodiments, 802.11ah can support meter type control / machine type communication, such as MTC devices in a macro coverage area. MTC devices can have certain capabilities, e.g., limited capabilities, including support for (e.g., support only) certain bandwidths and / or limited bandwidth. MTC devices can include a battery with a battery life above a threshold (e.g., to maintain a very long battery life).

[0059] WLAN systems that can support multiple channels and channel bandwidths such as 802.11η, 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 a STA from among all STAs operating in the BSS that supports the smallest bandwidth operating mode. In the example of 802.11ah, for a STA (e.g., MTC-type device) that supports (e.g., only supports) a 1 MHz mode, the primary channel can be 1 MHz wide, even if other STAs in the AP and 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) settings can depend on the status of the primary channel. If the primary channel is busy, e.g., due to a STA (that only supports a 1 MHz operating mode) transmitting to the AP, then the entire available frequency band can be considered busy, even if most of the frequency band remains idle and can be available.

[0060] In the United States, the available frequency bands for 802.11ah can be from 902 MHz to 928 MHz. In Korea, the available frequency bands can be from 917.5 MHz to 923.5 MHz. In Japan, the available frequency bands can be from 916.5 MHz to 927.5 MHz. The total bandwidth available for 802.11ah can be 6 MHz to 26 MHz, depending on the country code.

[0061] Figure 1D is a system diagram illustrating the RAN 113 and the CN 115 according to an embodiment. As

[0062] The RAN 113 can include gNBs 180a, 180b, 180c, although it will be appreciated that the RAN 113 can include any number of gNBs while remaining consistent with an embodiment. The gNBs 180a, 180b, 180c can each include one or more transceivers for communicating with the WTRUs 102a, 102b, 102c over the air interface 116. In one embodiment, the gNBs 180a, 180b, 180c can implement MIMO technology. For example, gNBs 180a, 108b can utilize beamforming to transmit signals to and / or receive signals from the gNBs 180a, 180b, 180c. Thus, the gNB 180a, for example, can use multiple antennas to transmit wireless signals to, and / or receive wireless signals from, the WTRU 102a. In an embodiment, the gNBs 180a, 180b, 180c can implement carrier aggregation technology. For example, the gNB 180a can transmit multiple component carriers (not shown) to the WTRU 102a. A subset of these component carriers can be on unlicensed spectrum while the remaining component carriers can be on licensed spectrum. In an embodiment, the gNBs 180a, 180b, 180c can implement Coordinated Multi-Point (CoMP) technology. For example, WTRU 102a can receive coordinated transmissions from gNB 180a and gNB 180b (and / or gNB 180c).

[0063] The WTRUs 102a, 102b, 102c can communicate with gNBs 180a, 180b, 180c using transmissions associated with a scalable numerology. For example, OFDM symbol spacing and / or OFDM subcarrier spacing can vary from transmission to transmission, from cell to cell, and / or from portion of the wireless transmission spectrum to portion of the wireless transmission spectrum. The WTRUs 102a, 102b, 102c can communicate with gNBs 180a, 180b, 180c using subframes or transmission time intervals (TTIs) of various or scalable lengths (e.g., containing varying number of OFDM symbols and / or lasting varying lengths of absolute time).

[0064] The gNBs 180a, 180b, 180c can be configured to communicate with the WTRUs 102a, 102b, 102c in a standalone configuration and / or a non-standalone configuration. In the standalone configuration, the WTRUs 102a, 102b, 102c can communicate with the gNBs 180a, 180b, 180c without also accessing other RANs, such as eNode-Bs 160a, 160b, 160c. In the standalone configuration, the WTRUs 102a, 102b, 102c can utilize signals according to one or more standards, such as UL / DL signals according to 5G NR standards. In the standalone configuration, the gNBs 180a, 180b, 180c can provide RAN integrity, RAN cell setup, and / or RAN cell modifications as well as small cell power control. In the standalone configuration, the WTRUs 102a, 102b, 102c can perform RAN measurements, such as reference signal received power (RSRP) measurements, reference signal received quality (RSRQ) measurements, reference signal time difference (RSTD) measurements, and / or the like. In the standalone configuration, the WTRUs 102a, 102b, 102c can also perform location measurements, such as Global Positioning System (GPS) measurements, assisted GPS measurements, and / or the like. In a non-standalone configuration, the WTRUs 102a, 102b, 102c can communicate with the gNBs 180a, 180b, 180c while also communicating with another RAN such as eNode-Bs 160a, 160b, 160c. For example, WTRUs 102a, 102b, 102c can implement DC principles to substantially simultaneously communicate with one or more gNBs 180a, 180b, 180c and one or more eNode-Bs 160a, 160b, 160c. In the non-standalone configuration, the eNode-Bs 160a, 160b, 160c can function as the mobile management entity (MME) to configure the WTRUs 102a, 102b, 102c in order to access and / or retain access to the gNBs 180a, 180b, 180c.

[0065] Each of the gNBs 180a, 180b, 180c can be associated with a particular cell (not shown) and can be configured to handle radio resource management decisions, handover decisions, scheduling of users in the UL and / or DL, support of network slicing, dual connectivity, interworking 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, and the like. As shown in FIG. 1C, the gNBs 180a, 180b, 180c can communicate with one another over an Xn interface. Figure 1D As shown in FIG. 1C, the gNBs 180a, 180b, 180c can also be configured to communicate with the core network 180 over an N2 interface.

[0066] Figure 1DThe CN 115, as shown in FIG. 10B, can 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 each of the foregoing elements are depicted as part of the CN 115, it will be appreciated that any of these elements can be owned and / or operated by an entity other than the CN operator.

[0067] The AMF 182a, 182b can be connected to one or more of the gNBs 180a, 180b, 180c in the RAN 113 via an N2 interface and can serve as a control node. For example, the AMF 182a, 182b can be responsible for authenticating the WTRUs 102a, 102b, 102c, supporting for network slicing (e.g., handling of different PDU sessions with different requirements), selecting a particular SMF 183a, 183b, management of the registration area, termination of NAS signaling, mobility management, and the like. Network slicing can be used by the AMF 182a, 182b in order to customize CN support for WTRUs 102a, 102b, 102c based on the type of service plans the WTRU 102a, 102b, 102c has subscribed to. For example, different network slices can be established for different use cases, such as services relying on ultra-reliable low latency (URLLC) access, services relying on enhanced massive mobile broadband (eMBB) access, services for machine type communication (MTC) access, and / or the like. The AMF 162 can provide control plane functionality for switching between the RAN 113 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.

[0068] The SMF 183a, 183b can be connected to AMF 182a, 182b in the CN 115 via an N11 interface. The SMF 183a, 183b can also be connected to UPF 184a, 184b in the CN 115 via an N4 interface. The SMF 183a, 183b can select and control the UPF 184a, 184b and configure the routing of traffic through the UPF 184a, 184b. The SMF 183a, 183b can perform other functions, such as managing and allocating UE IP address, managing PDU sessions, controlling policy enforcement and QoS, providing downlink data notifications, and the like. A PDU session type can be IP-based, non-IP based, Ethernet-based, and the like.

[0069] The UPF 184a, 184b can be connected to one or more of the gNBs 180a, 180b, 180c in the RAN 113 via an N3 interface, which can provide the WTRUs 102a, 102b, 102c with access to packet- switched networks, such as the Internet 110, to facilitate communications between the WTRUs 102a, 102b, 102c and IP-enabled devices. The UPF 184, 184b can perform other functions, such as routing and forwarding packets, enforcing user plane policies, supporting multi-homed PDU sessions, handling user plane QoS, buffering of downlink packets, providing mobility anchoring, and the like.

[0070] The CN 115 can facilitate communications with other networks. For example, the CN 115 can include, or can communicate with, an IP gateway (e.g., an IP multimedia subsystem (IMS) server) that serves as an interface between the CN 115 and the PSTN 108. In addition, the CN 115 can provide the WTRUs 102a, 102b, 102c with access to the other networks 112, which can include other wired and / or wireless networks that are owned and / or operated by other service providers. In one embodiment, the WTRUs 102a, 102b, 102c can be connected to a local DN 185a, 185b through the UPF 184a, 184b via the N3 interface between the UPF 184a, 184b and the UPF 184a, 184b and an N6 interface between the UPF 184a, 184b and the DN 185a, 185b.

[0071] In view of Figures 1A-1D And Figures 1A-1D In view of the corresponding description of the above, one or more or all of the functions described herein with reference 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, AMF 182a-b, UPF 184a-b, SMF 183a-b, DN 185a-b, and / or any other device(s) described herein can be performed by one or more emulation devices (not shown). The emulation devices can be one or more devices configured to emulate one or more or all of the functions described herein. For example, the emulation devices can be used to test other devices and / or to simulate network and / or WTRU functionality.

[0072] Simulation devices can be designed to perform one or more tests on other devices in a laboratory environment and / or a carrier network environment. For example, one or more simulation devices can perform one or more or all functions while being fully or partially implemented and / or deployed as part of a wired and / or wireless communication network to test other devices within the communication network. One or more simulation devices can perform one or more or all functions while being temporarily implemented / deployed as part of a wired and / or wireless communication network. Simulation devices can be directly coupled to another device for testing purposes and / or can be used to perform tests via over-the-air wireless communication.

[0073] One or more emulation devices can perform one or more (including all) functions without being implemented / deployed as part of a wired and / or wireless communication network. For example, emulation devices can be used in test scenarios within test laboratories and / or non-deployed (e.g., testing) wired and / or wireless communication networks to perform testing of one or more components. One or more emulation devices can be test rigs. Direct RF coupling and / or wireless communication via RF circuitry (e.g., which may include one or more antennas) can be used by the emulation devices to transmit and / or receive data.

[0074] The system can use a two-sided model (e.g., an AI / ML model for CSI feedback). In such a system, the features described herein can be implemented by the WTRU to determine and report CSI compression performance. The features described herein enable lifecycle monitoring (LCM) (e.g., monitoring the WTRU-side model, such as a machine learning (ML) model).

[0075] This article provides one or more features associated with model monitoring.

[0076] AI / ML frameworks for CSI compression can include two-sided models. In a two-sided model, CSI compression can be performed at the WTRU side. The compressed CSI can be fed back to the network (NW) and decompressed (e.g., restored) at the NW side. WTRU-side processing for CSI compression can include an ML encoder (e.g., possibly preceded by a preprocessing stage). NW-side processing can include an ML decoder (e.g., if preprocessing is performed at the WTRU, it may be followed by a postprocessing stage). The ML encoder and the corresponding ML decoder (e.g., their cooperative operation) can be referred to as an autoencoder (AE). Figure 2 A high-level block diagram of a CSI compression framework based on two-sided AI / ML is shown.

[0077] The ML encoder and ML decoder parts of the AE can be trained (e.g., separately or jointly) using a training dataset. The training of the ML model can be performed offline (e.g., before deploying the model at the node (WTRU and / or gNB)). During regular operation (e.g., at inference time), the performance of the compression (e.g., ML encoder) can degrade. Such degradation can occur, for example, if the distribution of the actual data does not match the distribution of the training dataset (this can be referred to as out-of-distribution operation). The WTRU or the network can detect when the performance of the CSI compression (and recovery) degrades. (e.g., after detection) appropriate mitigation mechanisms can be used. For example, mitigation can include model update, online training, or fallback to legacy reporting.

[0078] Monitoring the WTRU-side model can help the WTRU or the network detect and mitigate performance degradation (e.g., CSI compression performance degradation).

[0079] Machine learning based approaches (e.g., AE based approaches) can reduce the CSI feedback overhead. However, upon deployment of the ML model, the performance of the AI / ML based CSI compression can degrade (e.g., if the distribution (statistics) of the actual propagation channel is different from the distribution of the dataset used for training of the ML model). ML model monitoring can be used to detect whether the performance of the CSI compression degrades (e.g., so that appropriate mitigation measures can be applied).

[0080] The WTRU can have a copy of the ML decoder used by the NW. The WTRU can use this copy to determine the compression and reconstruction performance of the ML model monitoring. The WTRU can use the ML decoder to reconstruct the CSI (e.g., based on the compressed CSI reported back to the NW) and compare the reconstructed CSI with the CSI measured by the WTRU.

[0081] However, the availability of the ML decoder at the WTRU side can not be possible. Supporting multiple sizes can be required (e.g., due to different gNB Tx antenna port configurations, or supporting variable BW sizes). This can excessively increase the WTRU complexity (and can increase power consumption, as the WTRU can need to run the ML decoder separately to monitor the performance of the ML encoder).

[0082] The feature(s) described herein can relate to low complexity techniques for the WTRU to determine the compression performance for ML encoder performance monitoring.

[0083] Feature(s) are provided herein associated with measuring the distance between the estimated CSI and a set of reference vectors (e.g., to enable determining the compression performance).

[0084] Features associated with WTRU determining and reporting (e.g., to the NW) distance (e.g., best distance) are provided herein to accurately determine compression performance.

[0085] Features associated with WTRU determining and reporting how much distance is needed for compression performance determination are provided herein.

[0086] Features associated with WTRU determining a set of reference vectors for distance measurement are provided herein.

[0087] Given an assigned reporting size, the WTRU can determine (e.g., jointly determine) distance reporting parameters (e.g., such as quantization of distance, averaging) and the number of distances to report.

[0088] Features associated with detecting any mismatch between the reconstructed CSI (at the gNB) and the actual CSI measured at the WTRU using a set of reference vectors are provided herein.

[0089] For example, at the WTRU side, a first distance (or some other metric) between the actual CSI H j measured at the WTRU and the set of reference vectors can be measured. At the NW side, a second distance between the reconstructed CSI at the NW and the set of reference vectors (e.g., the same set of reference vectors) can be measured. If the two distances are found to be similar (e.g., then and the reconstruction of the compressed CSI at the gNB side can be acceptable (e.g., good). Otherwise, the AE-based reconstruction can be unacceptable (e.g., bad) and mitigation strategies (e.g., change or retraining of the encoder-decoder pair) can be employed.

[0090] A reference vector or basis vector can refer to a vector or tensor that is available at both the WTRU and the gNB / network. The reference vector or basis vector can be used as a reference. For example, a distance or metric or quantity can be evaluated based on a comparison with the reference vector or basis vector. The i-th reference vector can be denoted as R i .

[0091] A set of reference vectors or set of basis vectors can refer to a common set of multiple reference vectors, matrices, scalar values, or tensors that are available at both the WTRU and the gNB / network. If the reference is represented in matrix form, the set of reference vectors or set of basis vectors can be denoted as and if the reference is represented in vector form, as

[0092] ​The distance metric can be a function that maps an input tensor to a scalar value. In this case, the distance associated with a measured CSI and a reference vector in the set of reference vectors can be a scalar value. For example, the distance metric (e.g., distance dist(H j , R i ) can refer to any function that maps a pair of vectors or tensors to a scalar. Such a function can be referred to as a distance or metric, but can not adhere to the rules associated with (e.g., required to be) a distance or metric. The terms distance, distance metric, or metric can be used interchangeably herein.

[0093] The terms original channel, full channel (FC), full H, and full channel response matrix can be used interchangeably herein.

[0094] The feature(s) described herein can reduce WTRU complexity (e.g., because such feature(s) can not involve a copy of the AI / ML decoder available at the WTRU).

[0095] The distance measurement(s) can involve minimal computation. The detection overhead (e.g., total computational overhead of detection) at the WTRU and gNB can be low.

[0096] The feature(s) described herein can introduce minimal additional transmission overhead. The feature(s) described herein can utilize sending statistical measurement values and quantized values to keep the transmission overhead low.

[0097] Feature(s) associated with detection performance are provided herein. Under the correct distance metric (e.g., for projection) and optimal set of reference vectors, the WTRU can be able to (e.g., guarantee) detect a degradation (e.g., any degradation) in compression performance up to (e.g., any) defined accuracy. As the number of distances evaluated (e.g., with different reference vectors) increases, the detection performance can improve.

[0098] The WTRU can be configured to perform distance measurements (e.g., to support performance determination by the NW side). The configuration can include the distance metric dist(.,.) to be used. The metric used for evaluation can be a function that maps a pair of input tensors (e.g., H j , R i) a function that maps to a scalar real value (e.g., any function). The function can not adhere to the rules / properties required to qualify a function as a "metric" or "distance" (e.g., such as the triangle inequality and symmetry). The choice of metric can depend on the application domain and / or the type of data. Depending on the application, the objectives can be different (e.g., the objective(s) can include low reconstruction error, high alignment or cosine similarity, high perceptual quality, etc.). A function that is suitable for the application and the objective(s) can be utilized.

[0099] For CSI feedback applications, a distance such as the L2 norm of the error, mean square error (MSE), or normalized mean square error (NMSE) can be relevant if the reconstruction quality is measured in terms of mean square error. Other norms of the reconstruction error can be used (e.g., the LI norm or the Lp norm, where p corresponds to any real value). If the reconstruction error vector is given by E, the Lp norm can be defined as:

[0100]

[0101] For CSI feedback use cases, if the reconstruction quality is measured in terms of the cosine similarity between the eigenvectors of the actual and reconstructed CSI, the generalized cosine similarity or squared generalized cosine similarity (SGCS) can be used as the distance function.

[0102] A projection can be used as a function (e.g., for reconstruction or compression). The projection can capture the inner product between (e.g., two) input values.

[0103] For video or imaging applications, a perceptual quality metric (e.g., structural similarity) can be used.

[0104] If the network identifies the best distance metric for the application at hand, the network can configure the WTRU with that metric. Potential options for the metric can be listed as a codebook. The network can indicate an index associated with the metric to be used (e.g., by the WTRU). Additional configurable parameters can be defined depending on the metric. For example (e.g., in the case of weighted MSE), the weights associated with each dimension can be configured (e.g., explicitly configured).

[0105] The configuration can include a reference set domain.

[0106] The reference set domain can indicate the domain of data used to evaluate the performance of the bilateral model. For example, the domain of data can be (e.g., directly) a tensor (e.g., a channel matrix for CSI or CSI compression use cases). The domain of data can be any derived quantity of a tensor (e.g., an eigenvector of a tensor, e.g., an eigenvector for CSI). Full channel (FC) or eigenvector (EV) can be used for CSI compression, or another (e.g., any other) quantity that can be derived from the CSI tensor can be used.

[0107] The configuration can include a dimension reduction across subbands.

[0108] CSI tensor H j may have dimensions of N c x N R x N T , where N c represents the number of subcarriers, N R represents the number of receive antennas, and N T represents the number of transmit antennas. To evaluate distances, a reference vector R i may have the same dimensions of N c x N R x N T , and can evaluate (e.g., a single) distance for each CSI. Given the high dimensionality of the data (and to evaluate more meaningful distances), R i with a smaller dimensionality (e.g., N R x N T ) can be used. In such cases, multiple distances can be evaluated (e.g., one distance per subcarrier, e.g., N c distances). The smaller dimensionality can be used by ignoring different dimensions (e.g., N c x N T or N T ) or by reducing the size of each dimension of R i (e.g., ).

[0109] Using a reduced dimensionality R i may result in multiple distances. For example, if R i has dimensions of N R x N T , N c different distances can be evaluated and transmitted. To reduce the number of distances to be transmitted, the network can configure the WTRU with different strategies that are specifically configured to reduce the number of distances.

[0110] Proximate distances with similar values can be combined. In such cases, if there are N c subcarriers, Nc If the distance is too large, then some subcarriers' distance can be combined to have a subband distance. The size and grouping of these subbands can be different from the size and grouping of the subbands used for the CSI reporting. The size and grouping of these subbands can be configured by the network (e.g., during a configuration phase).

[0111] To further reduce the values to be reported, various strategies can be employed. For example, the average and / or variance of the distances across the subbands can be evaluated. Moments (e.g., any other moment than the average, variance) can be utilized. Weighted versions of the moments (e.g., weighted average, weighted variance) can be utilized (e.g., because moments are statistical quantities and do not preserve location information). Functions (e.g., any function) that take a multi-dimensional vector and output a scalar can be used for such reduction.

[0112] Given a set of potential options for the dimensionality reduction function (e.g., average / variance / weighted average / weighted variance / max, etc.), the network can signal (e.g., explicitly signal) which function to use by the WTRU.

[0113] The configuration can include a reference set.

[0114] The reference vectors used for the distance evaluation can be the same at the network and the WTRU. One or more reference vectors can be defined (e.g., to ensure accurate detection). The maximum (e.g., required maximum) number of reference vectors can depend on the dimensionality of the data and the variability of the data to be encountered. For CSI compression use cases, H j The higher the variability of H j , the larger the number of reference vectors. The reference set can use orthogonal / orthonormal reference vectors (e.g., because such reference vectors can capture the most variability with the least number of vectors). The reference vectors can not be orthogonal.

[0115] The set of reference vectors can be a large set of CSI tensors H R (e.g., with N T N R N T N R N T The set of left EVs of the N R N T <T) can be used as the set of reference vectors.

[0116] The network can configure the WTRU with a set of reference vectors H The WTRU can be configured with a set of reference vectors (e.g., because the same set of reference vectors can be used for distance evaluation at both the WTRU and the network). The selection of the reference vectors can depend on the data encountered by the gNB. The gNB (e.g., each gNB) can select a different set of reference vectors.

[0117] The gNB / network can (e.g., select) to configure the WTRU with a set of reference vectors. The reference vectors can be ordered or unordered. An ordered set can relay information to the WTRU such as, for example, which reference vectors have higher importance and corresponding distances that can be prioritized.

[0118] Features associated with NW-side model performance evaluation based on reference vectors are provided herein.

[0119] A set of reference vectors (or sets) can be used to perform NW-size evaluation of the performance of a bilateral model.

[0120] At the WTRU-side, a first distance (or some other metric) between actual measurements (e.g., CSI H j ) at the WTRU and a set of reference vectors can be measured. Given a distance metric (dist(.,.)), a set of reference vectors , and an operational mode (FC or EV or the like), the WTRU can evaluate the distance associated with the input data.

[0121] For a CSI use case, the data can be a CSI tensor H c with dimensions N R N T N j and the reference vectors can have dimensions N R N T . The selected distance metric can be the L2 norm and the operational mode can be FC. In this case, N c distances (e.g., one distance per subcarrier) can be evaluated. In an example, N c may refer to the number of subbands (e.g., in the frequency domain) for which the channel response is measured. The distance (e.g., each distance) can be evaluated (for subcarrier / subband index nc) as:

[0122]

[0123] The N c evaluated distances can be combined based on a dimensionality reduction procedure configured by the WTRU. For example (e.g., if an average-based dimensionality reduction scheme is defined), an average of all N c distances can be computed and used to indicate the distance.

[0124] At the NW side, a second distance between the reconstructed measurements (e.g., CSI ) at the NW and the same set of reference vectors can be measured.

[0125] For the distance evaluation on the network side, the same parameters and procedures can be repeated. Instead of the actual CSI measured by the WTRU, the network / gNB can use the reconstructed measurements (e.g., estimated / predicted CSI, ) to evaluate the distance to the same set of reference vector(s).

[0126] If the two distances are found to be similar (e.g., ), then the reconstruction of the compressed CSI at the gNB side can be acceptable (e.g., good). Otherwise, the AE-based reconstruction can be unacceptable (e.g., bad). In this case, mitigation strategies (e.g., change or retraining of the encoder-decoder pair) can be employed.

[0127] Figure 3 An example of a set of reference vectors, measured CSI, and distance metrics is illustrated.

[0128] The WTRU can report the distance metrics and the selected reference vectors.

[0129] Example types of validation metrics are provided herein.

[0130] The WTRU can report information to enable lifecycle management, validation, or confirmation of the AI / ML model. The WTRU can report the results of the AI / ML model or the transport block. The WTRU can report validation metrics to enable the gNB to determine or validate the performance of at least one of the following: AI / ML model, AI / ML encoder, AI / ML decoder, transport block, receive block, and / or the like. The validation metrics reported by the WTRU can include a measured channel. For example, the WTRU can report channel measurements obtained from at least one received RS. The measured channel can include a total channel measurement or a set of feature values associated with the channel measurements.

[0131] The validation metrics reported by the WTRU can include a compressed channel. For example, the WTRU can report compressed channel measurements associated with (e.g., obtained from) the measured channel measurements.

[0132] The validation metrics reported by the WTRU can include a reference vector. For example, the WTRU can report one or more reference vectors (e.g., determined based on features described herein). The reference vector report can include an index of the reference vector (e.g., for each reported reference vector). The association between the index and the reference vector can be pre-configured or determined by the WTRU. The reference vector report can include parameters associated with the reference vector. For example, the parameters associated with the reference vector can include the reference vector (e.g., the reference vector itself), or a compressed version of the reference vector.

[0133] The validation metrics reported by the WTRU can include a distance metric. For example, the WTRU can report at least one distance metric. The distance metric can be determined based on at least one of: a measured channel, a compressed channel, a reference vector, a measurement value, or a configuration.

[0134] The validation metrics reported by the WTRU can include multiple reference vectors or distances. For example, the WTRU can report multiple reference vectors or distances (e.g., reference vectors or distances determined by the WTRU for a measured channel or a compressed channel).

[0135] The validation metrics reported by the WTRU can include an acknowledgment of reception of a distance metric. For example, the WTRU can receive one or more distance metrics from a node (e.g., gNB). The WTRU can acknowledge the reception of the distance metric. In an example, the WTRU can compare the received distance metric with a distance metric computed by the WTRU. The WTRU can report a difference between the received and computed distance metrics.

[0136] The validation metrics reported by the WTRU can include a request for a new set of reference vectors.

[0137] The validation metrics reported by the WTRU can include a request to train an AI / ML model. For example, the request to train an AI / ML model can include a model ID. The validation metrics reported by the WTRU can include an AI / ML model ID. The validation metrics reported by the WTRU can include an AI / ML encoder ID. The validation metrics reported by the WTRU can include a transport block ID.

[0138] Features associated with reporting resources for reporting validation metrics are provided herein.

[0139] The WTRU can report validation metric(s).

[0140] The WTRU can report the validation metric(s) using feedback resources configured for and for reporting the CSI. For example, the WTRU can report the validation metric(s) as part of a feedback report. In an example, the WTRU can report a set of reference vectors and / or a set of distance metrics (e.g., distance metrics associated with the set of reference vectors) as part of a feedback report including one or more compressed measurements. The feedback report resources can be configured to be periodic, for example.

[0141] The WTRU can report the validation metric(s) using feedback resources configured for reporting the validation metric(s). For example, the WTRU can have dedicated reporting resources for reporting the validation metric(s). The WTRU can be configured with or can be indicated of a relationship between a first and a second reporting resource. The validation metric(s) can be reported in the first reporting resource and the compressed measurement value(s) can be reported in the second reporting resource. The feedback report resources can be configured to be periodic, for example.

[0142] The WTRU can report the validation metric(s) using dynamic or aperiodic granted reporting resource(s). For example, the WTRU can be dynamically or aperiodically granted reporting resources to report the validation metric(s). In an example, the WTRU can request a dynamically granted reporting resource to transmit the validation metric(s).

[0143] The WTRU can report the validation metric(s) using semi-persistent feedback resource(s). For example, the WTRU can be triggered to feedback the validation metric(s) using semi-persistent resources. The trigger to use semi-persistent resources can include at least one of the following: an indication from a gNB, a determination that a measurement value is above or below a threshold, AI / ML model performance degradation, QoS degradation, and / or ACK-NACK performance degradation.

[0144] If the WTRU reports a subset of the validation metric(s) in a resource that does not include an associated measurement value or an associated other validation metric, the WTRU can be configured with a relationship between reporting resources (e.g., different reporting resources). The relationship between reporting resources can depend on the type of validation metric reported in the reporting resource. For example, the WTRU can be configured with a first reporting resource that can include a set of reference vectors. The WTRU can determine (e.g., and report) one or more distance metrics in a reporting resource as a function of previously reported reference vectors (e.g., a most recently reported set of reference vectors). The WTRU can determine (e.g., and report) one or more distance metrics in a reporting resource as a function of previously confirmed reference vectors (e.g., where the confirmation is from another node such as a gNB). The reporting resource with the one or more distance metrics can be associated with at least one reporting resource with one or more measurement values or compressed measurement values. A relationship between the resource including the measurement values and the resource including the distance metrics can be configured.

[0145] A priority can be assigned or configured for the validation metric report. The WTRU can determine whether to report the validation metric based on (e.g., according to) at least one of the following: a validation metric report priority, a priority of other reports to be reported in the resource, and / or a payload of the feedback resource. The validation metric priority can be determined by at least one of the following: a priority of the associated data, a priority of the associated feedback report (e.g., measurement value), a priority of the function to be validated, and / or a value of the validation metric. The validation metric priority can be determined based on (e.g., according to) the value of the validation metric, e.g., the validation metric priority can be set based on determining that the distance metric is greater than a threshold. In an example, the WTRU can compare the WTRU determined distance metric to the gNB determined distance metric. If the difference is greater or less than a threshold, the WTRU can increase or decrease the priority of the validation metric feedback report.

[0146] Provided herein are feature(s) associated with a trigger and a keep-out distance.

[0147] The WTRU can be configured to report the validation metric periodically, aperiodically, or semi-persistently. For example, the WTRU can be configured to report the validation metric in the same resource as the associated feedback report. In another example, the WTRU can be configured with periodic resources to report the validation metric. The periodic resources can be associated with the periodic resources used to report the associated feedback report.

[0148] The WTRU can be triggered to determine or maintain the validation metric, or to report or start reporting the validation metric. For example, the WTRU can be triggered by a measurement value. For example, the WTRU can perform a measurement (e.g., RSRP, RSSI, RSRQ, CO, RI, CQI, PMI, LI, SINR, Doppler shift, Doppler spread, mean delay, delay spread, AoA, AoD, etc.) on an RS. Based on the value being above or below a threshold, the WTRU can be triggered to determine, maintain, report, and / or start reporting the validation metric. In an example, the WTRU can determine a value of the validation metric (e.g., a value of one or more distance metrics associated with one or more reference vectors). Based on the validation metric being above or below a threshold, the WTRU can be triggered to determine, maintain, report, and / or start reporting the validation metric. The WTRU can be triggered to determine, maintain, report, and / or start reporting based on a change in the validation metric. For example, if the validation metric changes by more than a threshold compared to a previously measured or reported validation metric, the WTRU can be triggered to report the validation metric (and possibly the previous validation metric).

[0149] A WTRU can be triggered based on a result of an associated function. For example, a WTRU can be triggered to determine, maintain, report, and / or start reporting a verification metric based on at least one of the following: a feedback reporting timing, a beam failure detection, a radio link failure, an unlicensed channel access result (e.g., successful LBT or unsuccessful LBT), an ACK / NACK transmission, an SRS transmission, an uplink control information (UCI) transmission, an RS reception, a PDCCH or PDSCH reception, a paging message reception, a SIB reception, a RACH procedure, an AI / ML model change, a determination of one or more NACKs, a reception of a retransmission grant, and / or the like.

[0150] A WTRU can be triggered based on a timing. For example, a WTRU can be triggered based on an absolute time, a slot, or a frame to determine, maintain, report, and / or start reporting a verification metric. In another example, a WTRU can be triggered based on a relative time, a slot, or a frame (e.g., relative to another event) to determine, maintain, report, and / or start reporting a verification metric. The relative timing can be relative to an event such as at least one of the following: a feedback reporting timing, a beam failure detection, a radio link failure, an unlicensed channel access result (e.g., successful LBT or unsuccessful LBT), an ACK / NACK transmission, an SRS transmission, a UCI transmission, an RS reception, a PDCCH or PDSCH reception, a paging message reception, a SIB reception, a RACH procedure, an AI / ML model change. In another example, a WTRU can be triggered based on a time since a previous reporting of a verification metric to determine, maintain, report, and / or start reporting a verification metric. In another example, a WTRU can be triggered based on a time period since a triggering condition was first met to determine, maintain, report, and / or start reporting a verification metric.

[0151] A WTRU can be triggered based on a transmission performance. For example, a WTRU can be triggered based on a performance of a UL or DL or SL transmission to determine, maintain, report, and / or start reporting a verification metric. For example, a WTRU can be triggered based on an ACK / NACK performance (e.g., the triggering can depend on whether a percentage of NACKs over a certain time period is greater or less than a threshold).

[0152] A WTRU can be triggered based on a QoS or an instantaneous QoS.

[0153] A WTRU can be triggered by receiving a request from another node. For example, a WTRU can be triggered to determine, maintain, report, and / or start reporting a verification metric based on receiving an indication from a gNB. The indication can be at least one of the following: a PDCCH indication, a DCI indication, a MAC CE indication, an RRC (re)configuration, a DL RS reception, a SL RS reception, a RAR message, and / or the like.

[0154] A WTRU can be triggered by a change in a scenario. For example, a WTRU can be triggered to determine, maintain, report, and / or start reporting a validation metric based on a change in a scenario, e.g., such as changing from being in a fixed location to being mobile. A WTRU can be triggered by at least one of a change in mobility, a change in LOS / NLOS, a change in a measured value that is greater than a threshold value (e.g., RSRP, RSSI, RSRQ, CO, RI, CQI, PMI, LI, SINR, Doppler shift, Doppler spread, average delay, delay spread, AoA, AoD, etc.), a change in a beam (e.g., Rx beam, Tx beam, or Tx / Rx beam pair), a change in a cell, a change in a TRP, a change in QoS (e.g., a new traffic type associated with new requirements), etc. Such a change in a scenario can be associated with a change in an AI / ML model.

[0155] A WTRU can be triggered based on a previous trigger. For example, a WTRU can be triggered to determine, maintain, report, and / or start reporting a validation metric based on whether a WTRU was previously triggered to measure, maintain, report, and / or start reporting a validation metric, and / or a timing of such a previous trigger, and / or a type of previous trigger.

[0156] A WTRU can be triggered based on receiving a validation metric report from another node. For example, a WTRU can be triggered to determine, maintain, report, and / or start reporting a validation metric based on receiving a validation metric computed by a gNB. For example, a WTRU can receive one or more distance metrics computed by a gNB. A WTRU can be triggered to report a validation metric based on at least one of receiving a validation metric computed by a gNB, a value of a validation metric computed by a gNB, a difference between a validation metric computed by a gNB and a validation metric computed by an associated WTRU, etc.

[0157] A WTRU can be triggered based on a counter or a timer. For example, a WTRU can be configured with a trigger counter and / or timer. If a number of trigger events is greater than a configurable value, a WTRU can be triggered to determine, maintain, report, and / or start reporting a validation metric, possibly during a configured amount of time. A trigger event can be any of the triggers described herein.

[0158] A WTRU can be triggered to stop reporting a validation metric using any of the triggers described herein.

[0159] A WTRU can determine content of a validation metric report based on a configuration or based on any of the triggers described herein.

[0160] Features are provided herein associated with WTRU measurement and reporting of distance / metric for performance monitoring of AI / ML models.

[0161] Example configuration aspects are provided herein. Example reference set configuration(s) are provided herein.

[0162] A WTRU can receive configuration information. The configuration information can indicate a set of reference vectors, input data type, and distance metric type. For example, a WTRU can be configured with multiple reference sets (e.g., I reference sets) to support monitoring the performance (e.g., compression) of a bilateral AI / ML model. A reference set can be defined as where i = {1,..., I} and K max may represent the maximum number of candidates associated with a particular reference set. K max A WTRU can change from one reference set to another reference set (e.g., based on NW configuration). A candidate (e.g., each candidate) in a reference set can represent a scalar (e.g., eigenvalue), a vector (e.g., principal eigenvector), a matrix (e.g., wideband channel matrix or multiple eigenvectors in a wideband), or a 3D array or tensor (e.g., multiple channel matrices or eigenvectors across multiple subbands). A WTRU can be configured or indicated with a particular reference set (e.g., selected by a NW) to perform and report measurement values (e.g., distance) for AI / ML model monitoring.

[0163] A reference set (e.g., each of the reference sets) can be characterized and indicated with one or more parameters. For example, a reference set (e.g., each of the reference sets) can be characterized and indicated with a ReferenceSet-ID. The ReferenceSet-ID can represent a reference set logical identity. The reference set logical identity can indicate (e.g., to a WTRU) the reference set to be used for distance measurement.

[0164] A reference set (e.g., each of the reference sets) can be characterized and indicated with a ReferenceSet-MaxSize(K max ). The ReferenceSet-MaxSize(K max ) can indicate the maximum number of candidates (e.g., vectors, matrices, tensors) that a WTRU can use from a configured reference set.

[0165] A reference set (e.g., each of the reference sets) can be characterized and indicated with a ReferenceSet-NrofdistanceToreport(k). The ReferenceSet-NrofdistanceToreport(k) can indicate the number of configured K maxThe number of K distances to report out of the K candidates in the reference set. If (e.g., only if) k is less than K max ReferenceSet-NrofdistanceToreport(k) can be signaled. Otherwise, if ReferenceSet-NrofdistanceToreport(k) is not signaled, the WTRU can assume that k is equal to K max .

[0166] A reference set (e.g., each of the reference sets) can be characterized and indicated with ReferenceSet-K max -mode. ReferenceSet-K max -mode can be a binary parameter that indicates whether the configured K max candidates are ordered (e.g., sorted). For example, if ReferenceSet-K max -mode is set to “1”, the configured set can be ordered. In this case, the WTRU can assume that k distance measurements will be made on the first k candidates in the configured reference set. Otherwise (e.g., if ReferenceSet-K max -mode is set to “0”), the WTRU can assume that the configured set is unordered. In this case, the WTRU can receive an indication of the indices of the candidates to use for distance measurements.

[0167] A reference set (e.g., each of the reference sets) can be characterized and indicated with ReferenceSet-k-locations. ReferenceSet-k-locations can indicate the indices of the k out of K max candidates to use for computing and reporting distance measurements. This parameter can be configured if (e.g., only if) the reference set candidates are unordered (e.g., the parameter ReferenceSet-K max -mode is set to “0”).

[0168] A reference set (e.g., each of the reference sets) can be characterized and indicated with ReferenceSet-Domain. ReferenceSet-Domain can indicate the domain (e.g., Eigen Vector (EV) or Full Channel) of the configured reference set, The WTRU can use the indicated domain to perform and report the measured distances. The WTRU can pre-process using the reference set domain before distance measurements (e.g., the reference set domain can be Full Channel, but the WTRU can receive an indication to measure distances in the EV domain).

[0169] A reference set (e.g., each of the reference sets) can be characterized and indicated with a ReferenceSet-DistanceDomain. The ReferenceSet-DistanceDomain can indicate the distance domain in which distance measurements are to be performed. For example, the ReferenceSet-DistanceDomain can indicate that the distance domain is a feature value, a feature vector, or a full channel.

[0170] A reference set (e.g., each of the reference sets) can be characterized and indicated with a ReferenceSet-DistanceMetricType. The ReferenceSet-DistanceMetricType can indicate the metric in which distance is to be measured. For example, the ReferenceSet-DistanceMetricType can include, but is not limited to, normalized mean square error (NMSE), (generalized) cosine similarity, matrix norm, projection, etc.

[0171] Provided herein are feature(s) associated with WTRU measuring distance.

[0172] A WTRU can be configured to perform and report k metrics (e.g., distance metrics / measurements) based on one or more of the following.

[0173] For example, a WTRU can be configured to perform and report k metrics (e.g., distance metrics / measurements) based on a configured reference set format (e.g., scalar, vector, matrix, tensor). A WTRU can be configured to perform and report k metrics (e.g., distance metrics / measurements) based on a configured distance metric type (e.g., NMSE, cosine similarity, projection). A WTRU can be configured to perform and report k metrics (e.g., distance metrics / measurements) based on a configured input data type (e.g., reference set domain, e.g., full channel, feature vector, feature value).

[0174] A WTRU can be configured to perform and report k metrics (e.g., distance metrics / measurements) based on a configured distance domain (e.g., full channel, feature vector, feature value).

[0175] The WTRU can use configured information to measure and report k distances. The configured reference set domain can differ from the configured distance domain. In this case, the WTRU can preprocess / transform the reference set domain to the distance domain based on the indicated metric (e.g., before calculating the distance). If a reference vector is associated with a first domain and the distance metric type is associated with a second domain, the WTRU can transform the reference vector from the first domain to the second domain. For example, the WTRU can be indicated or configured with a reference set in the full channel domain, while the configured distance domain is a feature vector or feature value. The WTRU can derive the feature vector or feature value associated with the reference set to measure the distance in the indicated or configured distance domain.

[0176] To perform k distance measurements, the WTRU can (e.g., it can first) estimate the distances across subbands (e.g., across all configurations of N). c Full channel matrix (each sub-band) For proper distance calculation, the WTRU can preprocess the estimated channel across subbands (e.g., all subbands).

[0177] WTRU can perform distance measurements based on a reference set with tensor candidates.

[0178] WTRU can be configured with a reference set. Each candidate in the reference set is a three-dimensional tensor format (e.g., Candidates (e.g., each candidate) can represent a feature vector in the full-channel domain or associated with the wideband full-channel domain. The WTRU can measure (e.g., directly measure) k distances based on a configured distance metric. For example, in the case of NMSE, the WTRU can compute the k distances as follows:

[0179]

[0180] Where R in This represents the nth sub-band associated with the i-th reference sample. F Let denote the Frobenius norm of the matrix, i = 1, ..., k.

[0181] The WTRU can be configured to report minimum and / or maximum distances across subbands (e.g., across all subbands). For example, the WTRU can compute the corresponding distance metric between the measured CSI and each reference vector in the set of reference vectors, and report either the minimum or maximum distance metric for that metric. In this case, k distances can be computed as follows:

[0182]

[0183] WTRU can perform distance measurements based on a reference set with matrix candidates.

[0184] WTRU can be configured with a reference set. The candidates in the reference set (e.g., each candidate) are in matrix format (e.g., Candidates (e.g., each candidate) can represent a wideband full-channel domain or an eigenvector associated with the wideband full-channel. For appropriate distance measurements, the WTRU can preprocess / transform the estimated channel to align with the reference set domain. For example, the WTRU can average the estimated channel across subbands (e.g., across all subbands) to compute the wideband channel. as follows:

[0185]

[0186] WTRU can be measured based on a configured distance metric and a configured distance domain. The WTRU calculates k distances between k indicated candidates from a configured reference set. For example, if the distance domain is indicated as feature vectors, the WTRU can (e.g., further) preprocess / transform the reference set candidates and the measured / estimated channel to the distance domain before calculating the distances. The k distances can be calculated based on the indicated distance metric type. The distance metric type can be NMSE. The WTRU can calculate the NMSE between the measured CSI and each (e.g., between) reference vectors in the set of reference vectors. For example, if the indicated distance metric type is NMSE, the WTRU can calculate the k distances as follows:

[0187]

[0188] Among them ||·|| F Let Frobenius norm be i = 1, ..., k.

[0189] The WTRU can receive the range domain as an indication of the feature vector. In this case, the WTRU can (e.g., first) derive the information from the wideband channel matrix. Associated feature vectors WTRU can compute distances in the EV domain. The distance metric type can be cosine similarity. WTRU can compute the cosine similarity between the measured CSI and (e.g., each) reference vectors in a set of reference vectors (e.g., between each other). For example, if the distance metric is indicated as cosine similarity, WTRU can compute k distances as follows:

[0190]

[0191] Where J represents the number of layers, and * represents the conjugate transpose. It is the channel matrix The j-th eigenvector, and rij is the j-th reference vector associated with the reference matrix R i .

[0192] The WTRU can perform distance measurements based on a reference set with vector candidates.

[0193] The WTRU can be configured with a reference set where the candidates (e.g., each candidate) in the reference set are in vector format (e.g., ). The candidates (e.g., each candidate) can represent a principal eigenvector associated with a wideband channel or can be a vector selected from a discrete Fourier transform (DFT) matrix. For proper distance measurements, the WTRU can pre-process / transform the estimated channel to be consistent with the reference set domain. For example, the WTRU can average the estimated channel across subbands (e.g., across all subbands) to compute a wideband channel The WTRU can derive a principal eigenvector The WTRU can compute k distances between the measured eigenvector and the indicated k references from the set based on the indicated distance metric (e.g., NMSE, cosine similarity).

[0194] The WTRU can perform monitoring based on eigenvalue measurements.

[0195] The WTRU can receive an indication that instructs the WTRU to measure and report one or more eigenvalues associated with the estimated wideband matrix, denoted as {a1, a2, …, a k}. The multiple eigenvalues can correspond to the number of layers supported by the WTRU.

[0196] The WTRU can be configured to compute and report the (e.g., one) eigenvalue of a configured subband (e.g., each configured subband). For n = 1, …, N c , the computed eigenvalue can correspond to the largest eigenvalue of the given subband channel matrix or the average across all eigenvalues associated with the given subband channel matrix H n . To monitor AI / ML performance, the NW can compute the set of eigenvalues of the decompressed channel , denoted as The NW can monitor performance based on a function of the reported eigenvalues associated with the encoder input and the eigenvalues derived based on the decoder output. For example, if the reported eigenvalues are close to the eigenvalues derived by the NW, the NW can conclude that the reconstruction (e.g., at the decoder output) is acceptable.

[0197] The WTRU can report the measured distances.

[0198] The WTRU can report k distances of measurements to be used for the bilateral AI / ML model monitoring. The WTRU can be configured to indicate the k distances of measurements and the output of the AI / ML encoder (e.g., compressed CSI). The WTRU can report the k distances in a MAC control element. The WTRU can report the k distances on a PUCCH resource. The WTRU can report the k distances in a PUSCH resource.

[0199] The WTRU can be configured to report the k distances periodically, semi-persistently, or aperiodically. The aperiodic reporting can occur based on one or more triggers. The WTRU can be configured to transmit the k distances based on one or more preconfigured events. For example, if the performance of the PDSCH changes (e.g., BLER exceeds a configured threshold), the WTRU can be configured to transmit the k distances. In another example, if the input data drift is identified by the WTRU, the WTRU can be configured to transmit the k distances. The WTRU can be configured to transmit the k distances if the AI / ML mode is switched.

[0200] Features are provided herein associated with WTRU measurements and reporting of distances and / or metrics for performance monitoring.

[0201] The WTRU can measure and report k distances between the measured CSI and the k reference vectors. This can enable the NW side to determine the compression performance (e.g., for life cycle management).

[0202] A WTRU in a system using a bilateral model (e.g., AI / ML model) for compression (e.g., for CSI feedback) can be configured to perform distance measurements (e.g., to support NW side compression performance determination). The configuration can include one or more of the following: a distance metric type to be used (e.g., NMSE, cosine similarity, matrix norm, projection, etc.); a set of reference vectors for distance measurement (e.g., K max total reference vectors in the set); an input data type (e.g., full channel response matrix (full H) measured at the WTRU side, or eigenvectors (EV) of the channel response); and / or a number k of distances to be reported (e.g., where k is less than or equal to the size K max of the reference set).

[0203] A WTRU can receive a CSI-RS. The WTRU can determine a measured CSI based on measurements associated with the CSI-RS. The WTRU can generate a compressed CSI from the measured CSI (e.g., based on the measured CSI and a configured input data type (full H or EV)). The WTRU can calculate a number k of distance metrics (e.g., between the measured CSI and each of a set of reference vectors) (e.g., based on a configured distance metric type). The WTRU can send a report to a network node. The report can indicate the compressed CSI and / or the determined k distance(s) between the measured CSI and the reference vector(s). max A device (e.g., a wireless transmit / receive unit (WTRU)) can be configured to receive configuration information. The configuration information can indicate an input data type, a set of reference vectors, and a distance metric type. The device can receive a channel state information (CSI) reference signal (CSI-RS). The device can determine a measured CSI based on measurements associated with the CSI-RS. The device can generate a compressed CSI based on the measured CSI and the input data type. The device can calculate (e.g., based on the distance metric type) a distance metric (e.g., a distance between) associated with the measured CSI and a reference vector of the set of reference vectors. The device can send a report to a network node. The report can include the compressed CSI and the distance metric associated with the measured CSI and the reference vector of the set of reference vectors.

[0204] A device (e.g., a wireless transmit / receive unit (WTRU)) can be configured to receive configuration information. The configuration information can indicate an input data type, a set of reference vectors, and a distance metric type. The device can receive a channel state information (CSI) reference signal (CSI-RS). The device can determine a measured CSI based on measurements associated with the CSI-RS. The device can generate a compressed CSI based on the measured CSI and the input data type. The device can calculate (e.g., based on the distance metric type) a distance metric (e.g., a distance between) associated with the measured CSI and a reference vector of the set of reference vectors. The device can send a report to a network node. The report can include the compressed CSI and the distance metric associated with the measured CSI and the reference vector of the set of reference vectors.

[0205] The input data type can be a full channel matrix or an eigenvector. The device being configured to calculate a distance metric associated with the measured CSI and a reference vector of the set of reference vectors can involve the device being configured to calculate a distance metric between the measured CSI and each reference vector of the set of reference vectors.

[0206] A WTRU can determine a number of distances / metrics to report.

[0207] Example configuration aspects are provided herein.

[0208] The WTRU can receive (e.g., be configured to receive) a CSI-RS for CSI estimation. The WTRU can determine the measured CSI based on the measurement associated with the CSI-RS. The WTRU can generate / compute a compressed CSI based on the input data type and the measured CSI. The input data type can include (but is not limited to) full channel, feature vectors, etc. The WTRU can be configured to report distance metrics with respect to the channel (e.g., to enable the detection and / or mitigation of errors in channel reconstruction). The WTRU can receive (e.g., from a gNB) configuration information (e.g., one or more configuration aspects) associated with the determination of multiple distance metrics to be reported. For example, the configuration information can indicate the input data type, the set of reference vectors, the distance metric type, and the error threshold. The WTRU can be configured to feed back distance metrics (e.g., to ensure that the number of distance metrics in the feedback meets pre-configured criteria).

[0209] WTRU can be configured with one or more reference vectors {R1, R2…R}. n This is used for distance metric feedback. The reference vector can correspond to a fundamental vector across the entire channel space. The reference vector can be predefined. The reference vector can be cell-specific and / or configured in broadcast signaling. The reference vector can be configured in RRC signaling. The reference vector configuration can be WTRU-specific. The reference vector can correspond to an feature vector. The reference vector can correspond to a DFT vector. The reference vector can be a channel matrix, channel implementation / sample, etc.

[0210] The WTRU can be configured to determine / derive a set of distance metrics (e.g., based on the configured distance metric type). A distance metric can refer to the distance between a channel measured at the WTRU and one or more reference vectors configured for the WTRU. Distance metrics can be configured such that the quality of the reconstruction can be inferred based on the distance relationships between the reference vectors and the measured channel on the WTRU and the reconstructed channel at the gNB. For example, the set of distance metrics may include a first distance metric associated with a first reference vector in the set of measured CSIs and reference vectors, and a second distance metric associated with a second reference vector in the set of measured CSIs and reference vectors, etc. For example, the channel matrix H measured at the WTRU and the channel reconstructed at the gNB... The relationships between them can be based on H and {R1, R2…R}. n The distance between (one or more) of} and {R1, R2…R n Distance metrics are used to express / infer / derive / determine the distances between (one or more) channels H and H. Distance metrics can (e.g., explicitly or implicitly) indicate performance metrics associated with CSI compression. For example, a distance metric can be used to measure the distances between two channels H and H. proxies for similarity between.

[0211] The distance metric can be configured as a cosine similarity metric or a variation thereof. For example, the WTRU can determine the set of distances by computing the cosine similarity of the measured CSI and the first reference vector and the cosine similarity of the measured CSI and the second reference vector. The distance metric can be a normalized mean square error. For example, the WTRU can determine the set of distances by computing the NMSE of the measured CSI and the first reference vector and the NMSE of the measured CSI and the second reference vector. The distance metric can be a Euclidean distance. The distance metric can be a projection of the channel matrix on the reference vectors.

[0212] The WTRU can be preconfigured with one or more distance metric types. The WTRU can derive the type of distance metric based on preconfigured rules. For example, the WTRU can determine the type of distance metric to use based on the feedback overhead / payload size, the number of reference vectors, the subband size, the error bound, the type of input used for compression (e.g., full channel or eigenvectors), and / or the like. The WTRU can be configured with a percentage error threshold Thr err . The WTRU can use the percentage error threshold to determine the number of distance metrics to include in the feedback. For example, the WTRU can determine a subset of the set of distance metrics (e.g., based on the error threshold). The set size of the subset can indicate the number of distance metrics in the subset. Different Thr err values can be configured for the WTRU. The WTRU can select the threshold to use based on the type of distance metric, the feedback overhead / payload size, the type of input used for compression, and / or the like.

[0213] The WTRU can determine the number of distance metrics to report (e.g., the number of distance metrics in the subset of distance metrics).

[0214] The WTRU can be configured to determine the number of minimum distances k to feedback according to a preconfigured condition. For example, the WTRU can be configured to determine and report k distances based on (e.g., according to) the configured error threshold. For example, the WTRU can compute the distance between each of the configured reference vectors and the measured channel or a preprocessed version thereof (e.g., eigenvectors).

[0215] The WTRU can determine the set of distance metrics by determining the number of minimum distances from the set of distances for which the energy associated with the distance is greater than a threshold percentage of the total energy. The threshold percentage can be equal to one hundred minus the error threshold (e.g., “(100-Thr errFor example, the WTRU can determine a minimum number of distances k such that the energy of the distances (e.g., projections) with k reference vectors is greater than a (100-Thr err ) percent of the total energy (e.g., such that the reported distances save at least a (100-Thr err ) percent of the energy).

[0216] The WTRU can send a report to the network node. The report can indicate the compressed CSI and the number of distance metrics in the subset. For example, the WTRU can be configured to report a minimum / default number of distance metrics (e.g., k min ) and a maximum number of distance metrics (e.g., K max ). The WTRU can select the number of distance metrics k actual (e.g., dynamically) based on one or more criteria described herein. For example, the WTRU can be configured to select a minimum number of distances (e.g., in addition to the default number of distances) that satisfy a preconfigured criterion. The report can further indicate the distance values associated with the distance metrics in the subset of distance metrics. The report can indicate the indices associated with the reference vectors associated with the subset of distance metrics.

[0217] The WTRU can be configured with a set of default reference vectors. The distances of the default reference vectors can be included in the feedback report. The WTRU can be configured with a set of (e.g., optional) reference vectors. The distances of the optional reference vectors can be included in the feedback report (e.g., if the distances of the default reference vectors do not satisfy a configured error threshold). The WTRU can be configured with a priority associated with the reference vectors (e.g., associated with each of the reference vectors). The WTRU can be configured to include the distance metrics (e.g., starting from the highest priority reference vector to the lowest priority reference vector). The WTRU can be configured to determine the number of distance metrics to include in the report based on (e.g., as a function of) one or more of the following: the quantization applied to the distance metrics, the type of input (e.g., full channel or feature vector), the number of subbands, the overhead / payload size configured for the feedback, and the like.

[0218] A WTRU can determine the number of distance(s) to report based on the type and / or identity of the encoder and / or decoder configured for CSI compression. For example, a WTRU can be configured with a mapping between encoders or decoders or encoder-decoder pairs and distance report configurations. A distance report configuration can include the number of distance(s) to report, the type of distance metric to use, the type of quantization to apply, the set of reference vectors to use, and / or the like. An encoder (or decoder or encoder-decoder pair) for CSI compression can correspond (e.g., implicitly) to a distance feedback configuration. For example, if a WTRU detects a change in an encoder (or decoder or encoder-decoder pair), the WTRU can determine a change in the distance feedback configuration.

[0219] A WTRU can report distance measurement values.

[0220] A WTRU can be configured to report one or more distance metrics / parameters associated with performance monitoring. As used herein, such reporting can be referred to as distance metric feedback, distance metric reporting, and / or the like. Such performance monitoring can be applicable to a two-sided model. Such reporting can be periodic, semi-persistent, and / or event-triggered.

[0221] A WTRU can be configured to report distance metric(s) multiplexed with CSI feedback. For example, a WTRU can transmit CSI feedback in one or more (e.g., two) parts. A first part can include compressed CSI feedback. A second part can include distance metric(s). A WTRU can transmit CSI feedback in a first UCI. A WTRU can transmit distance metric feedback in a second UCI. A WTRU can transmit CSI feedback via UCI and distance metric(s) feedback via PUSCH. A WTRU can transmit distance metric feedback associated with multiple CSI feedbacks in (e.g., a single) UL PUSCH. A PUSCH resource can be a semi-persistent resource. A WTRU can transmit distance metric feedback in a MAC CE. A WTRU can be configured with a time and / or frequency relationship between CSI feedback and distance metric feedback. A WTRU can indicate (e.g., explicitly or implicitly) the relationship between CSI feedback and distance metric(s) feedback.

[0222] The WTRU can determine the number of distance metric(s) (e.g., k) to be transmitted. The WTRU can determine the number of distance metric(s) based on different parameters, as described herein. The WTRU can indicate (e.g., explicitly or implicitly) the number of distance metric(s) (e.g., k) included in the distance metric feedback. For a distance metric (e.g., each distance metric), the WTRU can indicate the reference vector with which the distance metric is associated. For example, the WTRU can indicate the logical identity of the reference vector for a distance metric (e.g., each distance metric). Such logical identity can be pre-configured for the WTRU.

[0223] The WTRU can be configured to report a minimum number of distance metrics (e.g., k min ) and a maximum number of distance metrics (e.g., K max ). The WTRU can then dynamically select the number of distance metrics k actual based on one or more criteria, as described herein. The WTRU can indicate the number of distance metrics k actual included in the distance metric feedback. The WTRU can indicate the number of additional distance metrics (e.g., k actual -k min ) included in the distance metric feedback. The WTRU can indicate (e.g., only indicate) the recommended number of distance metrics in the distance metric feedback. The WTRU can indicate the number of distance metrics included in the distance metric feedback, the actual distance, and the associated reference vector identity.

[0224] A WTRU can be configured to quantize distance metrics indicated in distance metric feedback. For example, a WTRU can be configured to apply quantization to distance metrics based on preconfigured quantization parameters and transmit quantized distances. For example, a WTRU can apply scalar or vector quantization to distance metrics. For example, a WTRU can apply uniform or non-uniform quantization to distance metrics. Quantization parameters can be configured by a network. One or more quantization parameters can be determined (e.g., dynamically determined) by a WTRU. A WTRU can determine quantization parameters based on a tradeoff between accuracy and overhead. For example, given a configured feedback resource (e.g., PUSCH, PUCCH, or the like), a WTRU can be configured to quantize distance metrics such that accuracy of CSI reconstruction is maximized. For example, a WTRU can apply different quantization profiles based on a feedback resource (e.g., PUSCH or PUCCH). For example, a WTRU can apply different quantization levels to different distance metrics (e.g., according to importance of distance metrics). A WTRU can indicate quantization parameters in distance metric feedback. A WTRU can be preconfigured with multiple quantization profiles. A quantization profile (e.g., each quantization profile) can include quantization configurations applicable to quantize distance metrics. For example, a quantization profile (e.g., each quantization profile) can be associated with a logical ID. A WTRU can select a quantization profile (e.g., one of the quantization profiles) and apply the quantization profile to distance metrics (e.g., all distance metrics) in distance metric feedback. A WTRU can select different quantization profiles specific to distance metrics (e.g., each distance metric). For example, a WTRU can quantize distance metrics with highest projected values with the largest number of bits. A WTRU can be configured to indicate quantization profile(s) applied to distance metric(s) by including a logical ID of a quantization profile in distance metric feedback.

[0225] A WTRU can determine a number of distances and / or metrics to report.

[0226] A WTRU can determine and report a number of reference vectors (k) for which to report distance metrics (e.g., according to a configured error threshold).

[0227] A WTRU in a system for CSI feedback using a double-sided model (e.g., an AI / ML model) can be configured to perform distance measurements (e.g., to support NW-side compression performance determination). The configuration can include: a distance metric type to use; a set of reference vectors; an input data type; and / or an error threshold percentage Thr err .

[0228] A WTRU can receive a CSI-RS. The WTRU can determine a measured CSI based on measurements associated with the CSI-RS. The WTRU can compute a compressed CSI from the measured CSI (e.g., based on a configured input data type (full H or EV)). The WTRU can determine how many distance metrics (k) to report (e.g., according to a configured error threshold). The WTRU can compute K max distance metrics between the measured CSI and a set of configured reference vectors. The WTRU can determine a minimum number of distance metrics k such that the energy of the distance metrics (e.g., projections) with the k reference vectors is greater than a (100-Thr err ) % of the total energy (e.g., the reported distance metrics save a (100-Thr err ) % of the energy).

[0229] The WTRU can report one or more of the following to the NW: the compressed CSI; the recommended number of distance metrics k; the values of the k distance metrics; and / or an indication (e.g., index) of the associated reference vectors.

[0230] The network can configure the WTRU to report the recommended number of distance metrics k.

[0231] A device (e.g., a wireless transmit / receive unit (WTRU)) can receive configuration information. The configuration information can indicate an input data type, a set of reference vectors (e.g., a first plurality of sets of reference vectors), a distance metric type, and an error threshold. The device can receive a channel state information (CSI) reference signal (CSI-RS). The device can determine a measured CSI based on measurements associated with the CSI-RS. The device can generate a compressed CSI based on the measured CSI and the input data type. The device can determine a number of distances to report (e.g., based on the error threshold). The device can transmit a report to a network node. The report can include the compressed CSI and the number of distances.

[0232] The configuration information can further indicate a set of reference vectors. The device can compute a plurality of distances, where the plurality of distances includes a distance between the measured CSI and each reference vector of the set of reference vectors. The device can determine a number of minimum distances from the plurality of distances for which an energy associated with the distance is greater than a threshold percentage of a total energy. The threshold percentage can be equal to one hundred minus the error threshold. The input data type can be a full channel matrix or an eigenvector.

[0233] The WTRU can determine a set of reference vectors and / or distances / metrics (e.g., best distances / metrics) to report.

[0234] Example configuration aspects are provided herein.

[0235] The WTRU can receive a first plurality of sets of reference vectors. For example, the gNB can identify Ml different sets of reference vectors, The reference vectors across different sets can be (e.g., can all be) different or partially overlapping (e.g., with a few common vectors). The reference vectors across different sets can be (e.g., can all be) the same (e.g., but with different ordering or priority associated with each of the base vectors).

[0236] The gNB can configure the WTRU with Ml different sets.

[0237] The WTRU can determine a second plurality of sets of reference vectors based on the measured CSI. For example, the WTRU can (e.g., independently, or after observing the Ml sets from the gNB) identify M2 (e.g., additional) sets of reference vectors

[0238] m e {1...M2}. The WTRU can determine that the M2 (e.g., additional) sets of reference vectors are more suitable for distance-based detection.

[0239] The WTRU can indicate the M2 different sets to the gNB. For example, both the WTRU and the gNB have access to the Ml + M2 sets.

[0240] The gNB can signal a parameter k (e.g., to indicate how many distances the WTRU is to initially report) to the WTRU. The gNB can configure the WTRU with an error threshold percentage Thr err for the reconstruction performance calculation (e.g., instead of defining the exact number of distances to report).

[0241] The WTRU can be configured (e.g., can receive the configuration information) via RRC signaling or via a control channel.

[0242] The WTRU can determine / select a set of reference vectors.

[0243] As described herein, for a set (e.g., each of the Ml + M2 sets), the WTRU can evaluate K max distances.

[0244] The WTRU can select a set of reference vectors from a set of combinations of sets of reference vectors (e.g., where the set of combinations of sets of reference vectors includes the first plurality of sets of reference vectors and the second plurality of sets of reference vectors) based on the distance metric type. For example, if the WTRU is configured with k (e.g., distances to report), the WTRU can (e.g., jointly) evaluate the best set (outside of the Ml + M2 sets) and the best k distances to report from the best set.

[0245] The "best distance" can depend on the distance metric used. For example, the distance metric type can be a projection metric. If projection is used as the distance metric, the best reference vector can be determined based on the projection of the CSI on the largest normalized reference vector. For example, the WTRU can select the set of reference vectors for which the projection of the measured CSI on the normalized reference vectors in the set of reference vectors is the largest. If MSE or L2 norm is the distance metric being used, the reference vector for which the magnitude of the distance is the smallest can be more useful (e.g., can be considered the best distance). For example, the WTRU can select the set of reference vectors associated with the smallest mean squared error between the measured CSI and the reference vectors in the set of reference vectors. The distances corresponding to the k best reference vectors can be the best k distances.

[0246] The set (e.g., for which the best k distances collectively represent the most of the energy of the input CSI) can be used for reporting. The WTRU can select the set of reference vectors that saves the most total energy. For example, in the case where projection is the distance metric of choice, the set for which the best k reference vectors represent the most of the total energy of the input CSI (e.g., the normalized sum of the projections if the reference vectors are orthogonal) can be the best set for distance reporting.

[0247] The configuration information can indicate an error threshold percentage. The error threshold percentage can be in a range between 0 and 100% (e.g., where 0% indicates no error). The WTRU can select one or more reference vectors from the selected set of reference vectors based on the error threshold percentage. In this case, the report can indicate the selected one or more reference vectors. For example, if the WTRU is configured with Thr err (e.g., an error threshold percentage), the WTRU can evaluate the best set (e.g., outside of the M1+M2 sets), the value of k, and the best k distances reported from the best set.

[0248] For each of the sets (e.g., M1+M2 sets), the WTRU can determine the smallest number of distances k m such that the energy of the distances (e.g., projections) with the k m reference vectors is greater than (100-Thr err )% of the total energy (e.g., the selected distances save (100-Thr err )% of the energy).

[0249] The mthset with the smallest value of k m may be selected as the best set for distance-based reporting. The corresponding k m distances can be the best distances (e.g., k=k m ).

[0250] A WTRU can (e.g., dynamically) activate and / or deactivate a set of reference vectors. For example, if a WTRU receives an activation notification from the NW, the WTRU can activate a configured set of reference vectors. As another example, if a WTRU receives a notification to activate a new set, the WTRU can deactivate a previously configured reference set.

[0251] A WTRU can send a report to a network node. A WTRU can report a determined set of reference vectors. For example, the report can indicate a compressed CSI and at least a portion of the selected reference vector set.

[0252] A WTRU can report the identity of the selected set (e.g., out of the available M1+M2 sets). The set can be represented as a lookup table (LUT) or codebook (e.g., since the set is known at both the WTRU and gNB). An index (e.g., only an index) associated with the selected set can be reported by the WTRU.

[0253] If a WTRU is configured with Thr err A WTRU can report the value of k, indicating how many distances will be reported.

[0254] A WTRU can report the identity of the best k reference vectors (e.g., used for distance measurement).

[0255] A WTRU can report individual distances associated with the k reference vectors.

[0256] The selection and number of M2 sets specified by a WTRU can change over time. A WTRU can update the reference vectors in any M2 set (e.g., dynamically and / or aperiodically), or can introduce a new set of reference vectors.

[0257] A WTRU can determine a set of reference vectors and / or distances (e.g., best distances) and / or metrics to report.

[0258] A WTRU can select a set of reference vectors for which to determine and report distances (e.g., distances between measured CSI and selected set of reference vectors). A WTRU can determine k distances to report (e.g., best k distances) (e.g., based on channel conditions and / or preconfigured reporting parameters).

[0259] A WTRU in a system using a bilateral model (e.g., an AI / ML model) for CSI feedback can be configured to perform distance measurements. The network side can use the distance measurements to determine compression performance. The configuration can include: a distance metric type to use; an input data type; a set of multiple (M1) reference vectors. For example, the set of reference vectors can have reference vectors that are different from other sets of reference vectors. The set of reference vectors can have the same reference vectors as other sets of reference vectors but with different ordering.

[0260] The WTRU can receive a CSI-RS. The WTRU can measure the CSI-RS to determine measured CSI. The WTRU can determine an additional (M2) set of reference vectors (e.g., based on the measured CSI). The WTRU can indicate the additional M2 set of reference vectors to the network. The WTRU can compute compressed CSI from the measured CSI (e.g., based on the configured input data type (full channel (H) or EV)).

[0261] For each set of reference vectors, the WTRU can determine the best k reference vectors to use. For example, the WTRU can select the vectors that correspond to the k smallest distances (or k largest projections).

[0262] The WTRU can determine which set of reference vectors to use (e.g., from the configured M1 and additional M2 sets of reference vectors) (e.g., determined according to a reporting configuration). For example, if the WTRU is configured to report k distances, the WTRU can select the reference set with the smallest distances that correspond to the best k reference vectors in the set. If the WTRU is configured with an error threshold percentage Thr err for reconstruction performance, the WTRU can select the reference set that saves the most energy (e.g., the highest total energy of all reference vectors in the reference set).

[0263] The WTRU can report one or more of: the compressed CSI; the selected set of reference vectors; an indication (e.g., an index) of the determined best reference vectors (e.g., associated with each reported distance / metric); and / or the determined best k distances / metrics between the measured CSI and the reference vectors.

[0264] A device (e.g., a wireless transmit / receive unit (WTRU)) can receive configuration information. The configuration information can indicate an input data type, a first plurality of reference vector sets, a distance metric, and an error threshold. The device can receive a channel state information (CSI) reference signal (CSI-RS). The device can perform measurements on the CSI-RS to determine a measured CSI. The device can compute a compressed CSI based on the measured CSI and the input data type. The device can determine a second plurality of reference vector sets (e.g., based on the measured CSI). The device can select a reference vector set from the first plurality of reference vector sets and the second plurality of reference vector sets (e.g., based on the distance metric). The device can transmit a report to a network node. The report can include the compressed CSI, the selected reference vector set.

[0265] The device can transmit an indication of the second plurality of reference vector sets to the network node. The device can determine a subset of reference vectors for each reference vector set in the first plurality of reference vector sets (e.g., based on the distance metric and distances associated with the subset of reference vectors). The report can include the subset of reference vectors and an indication of the distances associated with the subset of reference vectors. Selecting the reference vector set from the first plurality of reference vector sets and the second plurality of reference vector sets can involve the device selecting the reference vector set with the best distance associated with the subset of reference vectors.

[0266] The configuration information can indicate an error threshold percentage. Selecting the reference vector set from the first plurality of reference vector sets and the second plurality of reference vector sets can involve the device selecting the reference set that saves the highest total energy for all reference vectors in the reference set (e.g., based on the error threshold percentage). The input data type can be a full channel matrix or an eigenvector.

[0267] Figure 4 An example of a WTRU determining a set of reference vectors is illustrated.

[0268] The WTRU can determine (e.g., jointly) distance / metric reporting parameters.

[0269] Example configuration aspects are provided herein.

[0270] A WTRU can receive (e.g., from a gNB) configuration information associated with a maximum payload size B (e.g., which can be received over a control channel (e.g., PDCCH)). The configuration can include an overhead threshold (e.g., a maximum overhead associated with transmissions of the WTRU to the gNB reporting distances). The WTRU can receive a distance metric type to use, a set of reference vectors, an error threshold, a similarity threshold, a quantization threshold, and / or an input data type. Based on the configuration information, the WTRU can determine (e.g., jointly) a number of k distances to report, subband grouping and averaging, and / or quantization per distance.

[0271] The WTRU can determine distance measurement parameters.

[0272] The WTRU can measure and evaluate distances. The WTRU can receive CSI-RS. The WTRU can determine measured CSI based on measurements associated with the received CSI-RS. The WTRU can generate / compute compressed CSI based on the measured CSI and input data type (e.g., full channel (FC) or eigenvector (EV)).

[0273] For a subband (e.g., each subband), the WTRU can estimate all K max distances / distance measures. The WTRU can decide which k < K max distance measures should be sent to satisfy a predefined error bound Thr err . For example, if projections are used as measures, the WTRU can determine the minimum k reference vectors such that the projection of the FC / EV on the k reference vectors saves (100-Thr err ) % of energy. The number of distances per subband can be affected by the maximum payload size B.

[0274] The WTRU can determine subband grouping and subbands of multiple groups (e.g., group subbands by distance) based on a similarity threshold (e.g., after computing distances for each subband). The grouping measure can be based on similarity (e.g., groups of subbands with similar distances are considered redundant and the redundancy can be avoided). Distances can not be similar. Distances can be close (e.g., very close). The WTRU can define a similarity threshold Thr sim as the maximum difference between reference distances (e.g., which can make two sets disjoint). Distances can be considered similar and associated subbands can be grouped in the same subband group if the difference is less than the threshold (e.g., Thr sim ). For example, the WTRU can determine a first distance measure between a first reference vector and a measured CSI. The WTRU can determine a second distance measure between a second reference vector and the measured CSI. The WTRU can determine a difference between the first distance measure and the second distance measure. In the case that the difference is less than the similarity threshold, the WTRU can group the subband associated with the first distance measure and the subband associated with the second distance measure in a subband group.

[0275] WTRU can optimize (e.g., jointly optimize) the overhead (e.g., instead of reporting the original distance values) (e.g., after grouping). WTRU can compute an average value associated with each group. Other moments can be used to distinguish groups (e.g., each group). Contribution weights can be computed for each distance (e.g., each of k distances). For example, the sum of the weights can equal 1. Subband grouping may be affected by the bit budget B.

[0276] The WTRU can quantize k distances for each subband group. The WTRU can determine the quantization level (e.g., based on a quantization threshold). Quantization can depend on bit budget constraints. The WTRU can use a uniform quantization function for each distance, or a mixture of uniform and non-uniform quantization functions. The quantization function can be dynamic and data-driven (e.g., designed to minimize quantization error). The number of quantization bits used for a distance (e.g., per distance) may be affected by the maximum payload size B. Using more quantization levels may increase the overhead and / or accuracy of compression performance detection. Similarly, using fewer quantization bits for a distance may decrease the accuracy of compression performance detection. The WTRU can use the quantization error Thr q The threshold is used to satisfy the maximum payload size B with acceptable quantization error.

[0277] For the configured reporting payload B (e.g., bit budget B), WTRU can determine the number (k) of distances to be reported, the quantization of the reported distances, and / or subband averaging.

[0278] WTRU can determine whether the predefined error bound Thr is met. err k <K max A distance.

[0279] WTRU can determine subband groupings. WTRU can determine groups that meet the similarity threshold Thr. sim The number of grouped subbands N G .

[0280] The quantization level indicates the number of bits used to represent the scalar distance value between reference vectors in the set of measured CSI and reference vectors. WTRU can determine whether the quantization error threshold Thr is met. q The quantization level Q (e.g., the number of bits used to represent a scalar distance value). WTRU can report one or more distance metrics from a set of distance metrics based on the number of bits.

[0281] WTRU can determine the total payload size based on the number of distances, the number of subbands in the group, and the quantization level. For example, WTRU can calculate the total payload size (e.g., by N...). P =k·N G• Q to compute).

[0282] If N p ≤ B, the WTRU can use k distance and quantization levels Q for N G subgroups.

[0283] If N p > B, the WTRU can change the threshold (e.g., such that Thr err = Thr err + A e , Thr sim = Thr sim + A s , and Thr q = Thr q + A q ). The WTRU can then (re)determine k < K err distances that satisfy the pre-defined error bound Thr max .

[0284] The WTRU can transmit the k distances per subband group with the determined quantization levels.

[0285] Figure 5 Examples of determining the number of distances to report (k), quantization of the reported distances, and subband averaging (e.g., grouping) for a configured reporting payload B (e.g., jointly) are illustrated.

[0286] The WTRU can report distance measurement parameters.

[0287] The WTRU can report the distance measurement parameters to the gNB using a control channel (e.g., PUCCH). The reported parameters can include one or more of the following: parameter k, k computed distances; subband grouping identification (e.g., a set of indices and subbands belonging to each index); average distance (e.g., or other moment) in each group; number of quantization bits for distances in a subband (e.g., each distance in each subband), and / or the like. If the quantization is not uniform, fields can be reserved to report the quantization parameters separately.

[0288] The WTRU can determine (e.g., jointly determine) the distance / distance metric and / or the reporting parameters.

[0289] The WTRU can determine (e.g., jointly determine) the number of distances (k) to report (e.g., based on the feedback reporting payload) for each subband, subband group, and / or distance measurement quantization (between the measured CSI and reference vector).

[0290] A WTRU in a system using a bilateral model (e.g., an AI / ML model) for CSI feedback can be configured to perform distance measurements (e.g., used by the NW side to determine compression performance). The configuration can include one or more of: a distance metric type to use; a set of reference vectors; an input data type; and / or a maximum overhead (B bits) associated with reporting of distance metrics.

[0291] A WTRU can receive a CSI-RS. The WTRU can determine measured CSI based on measurements of the CSI-RS. The WTRU can compute compressed CSI from the measured CSI (e.g., based on a configured input data type (full H or EV)). The WTRU can determine (e.g., jointly determine) a number of distances (k) to report, quantization of the reported distances, and / or subband averaging (e.g., according to a configured reporting payload B).

[0292] A WTRU can compute K max distances for an assigned subband (e.g., each assigned subband). The WTRU can determine which k distances to report for each subband (e.g., to meet a configured error threshold).

[0293] A WTRU can determine a grouping of subbands for averaging. The WTRU can determine quantization of distances (e.g., each of the k distances) for each subband.

[0294] A WTRU can report one or more of: the determined grouping of subbands for averaging; the measured k distances and indices of the corresponding reference vectors; and / or quantization information of the reported distances.

[0295] A device (e.g., a wireless transmission / reception unit (WTRU)) can receive configuration information. The configuration information can indicate an input data type, a set of reference vectors, a distance metric type, an error threshold, a similarity threshold, a quantization threshold, and an overhead threshold. The device can receive a channel state information (CSI) reference signal (CSI-RS). The device can determine measured CSI based on measurements associated with the CSI-RS. The device can generate compressed CSI based on the measured CSI and the input data type.

[0296] The device can determine the number of distances based on the error threshold. For example, the device can determine a set of distance metrics based on the distance metric type. A set size of the set of distance metrics can indicate a number of distance metrics in the set of distance metrics. The set of distance metrics can include a first distance metric associated with the measured CSI and a first reference vector of the set of reference vectors, and a second distance metric associated with the measured CSI and a second reference vector of the set of reference vectors. The device can determine the amount of the number of distance metrics to report based on the error threshold.

[0297] The device can determine the number of subband groupings and the number of grouped subbands based on the similarity threshold. The device can determine a quantization level based on the quantization threshold. The device can calculate a total payload size based on the number of distances, the number of grouped subbands, and the quantization level. The device can transmit a report to the network node. The report can indicate the compressed CSI.

[0298] In a case that the total payload size is less than or equal to the overhead threshold, the report can indicate the amount of the number of distance metrics, the number of grouped subbands, and the quantization level. The input data type can be a full channel matrix or a eigenvector.

[0299] The error threshold can be a first error threshold. The similarity threshold can be a first similarity threshold. The quantization threshold can be a first quantization threshold. In a case that the total payload size is greater than the overhead threshold, the device can increment the first error threshold to obtain a second error threshold; increment the first similarity threshold to obtain a second similarity threshold; increment the first quantization threshold to obtain a second quantization threshold; determine a second amount of the number of distance metrics based on the second error threshold; determine a second number of subband groupings and a second number of grouped subbands based on the second similarity threshold; determine a second quantization level based on the second quantization threshold; and calculate a second total payload size based on the second amount of the number of distance metrics, the second number of grouped subbands, and the second quantization level. The report can indicate the second amount of the number of distance metrics, the second number of grouped subbands, and the second quantization level.

[0300] Although the above-described features and elements are described in particular combinations, each feature or element can be used alone without the other features and elements or in various combinations with or without other features and elements.

[0301] While implementations described herein can consider 3GPP specific protocols, it is to be understood that the implementations described herein are not limited to such scenarios and can be applicable to other wireless systems. For example, while the solutions described herein consider LTE, LTE-A, New Radio (NR), or 5G specific protocols, it is to be understood that the solutions described herein are not limited to such scenarios and are applicable to other wireless systems as well. For example, while the systems have been described with reference to 3GPP, 5G, and / or NR network layers, contemplated embodiments extend beyond implementations using specific network layer technologies. Likewise, potential implementations extend to all types of service layer architectures, systems, and embodiments. The techniques described herein can be applied individually and / or in combination with other resource configuration techniques.

[0302] The processes described herein can be implemented in a computer program, software, and / or firmware incorporated in a computer- readable medium for execution by a computer and / or processor. Examples of computer-readable media include but are not limited to electronic signals (optical, electrical or electromagnetic) that are transmittable through a wired and / or wireless connections over the Internet and / or over a wired and / or wireless communication medium. Examples of computer-readable media include, but are not limited to, a random access memory (RAM), a read-only memory (ROM), a register, cache memory, semiconductor memory devices, magnetic media such as, but not limited to, internal hard disks and removable disks, magneto-optical media, and / or optical media such as compact disks (CDs) and / or digital versatile disks (DVDs). The processor associated with a software can be for implementing a radio frequency transceiver for use in a WTRU, terminal, base station, RNC, and / or any host computer.

[0303] It is to be understood that the entities performing the processes described herein can be logical entities that can be implemented in the form of software (e.g., computer-executable instructions) stored in the memory of a mobile device, network node, or computer system and executed by the processor of the mobile device, network node, or computer system. That is, the processes can be implemented in the form of software (e.g., computer-executable instructions) stored in the memory of a mobile device and / or network node, such as a node or computer system, that, when executed by the processor of the node, perform the processes in question. It is also to be understood that any transmitting and receiving processes illustrated in the figures can be performed by the communication circuitry of the node under the control of the processor of the node and the computer-executable instructions (e.g., software) executed thereby.

[0304] The various techniques described herein can be implemented in connection with hardware or software or, where appropriate, with a combination of both. Thus, the subject matter described herein can take the form of program code (e.g., instructions) embodied in tangible media such as a storage medium of a tangible medium of a machine-readable storage medium, wherein, when the program code is loaded into and executed by a machine such as a computer, the machine becomes an apparatus for practicing the subject matter described herein. In the case of program code execution on programmable devices as appropriate, the program code can execute individually or in a single processing unit or distributed across multiple processing units. Program code can be written in any form or language of the device's execution, including interpreted languages or, where desired, machine languages. In the case of implementation at least in part on a software platform or operating system on a computing device, the computing device generally includes a processor, a storage medium readable by the processor (including volatile or non-volatile memory and / or storage elements), at least one input device, and at least one output device. One or more programs can implement or utilize the processes described herein, for example, through the use of an API, reusable controls, or the like. Such programs are preferably implemented in a high level procedural or object oriented programming language to communicate with a computer system. However, the program(s) can be implemented in assembly or machine language, if desired. In any case, the language can be a compiled or interpreted language, and combined with hardware implementations.

[0305] Although example embodiments can refer to utilizing aspects of the subject matter described herein in the context of one or more stand-alone computing systems, the subject matter described herein is not so limited, but rather can be implemented in connection with any computing environment, such as a network or distributed computing environment. Still further, aspects of the subject matter described herein can be implemented across multiple processing chips or devices, and storage across multiple devices can be similarly affected. Such devices might include personal computers, network servers, handheld devices, supercomputers, or computers integrated into other systems such as automobiles and aircraft.

[0306] In describing the preferred embodiment of the subject matter of the present disclosure, specific terminology is employed for the sake of clarity. However, the claimed subject matter is not intended to be limited to the specific terms so selected, and it is to be understood that each specific element includes all technical equivalents that operate in a similar manner to accomplish a similar purpose.

Claims

1. A wireless transmit / receive unit (WTRU) comprising: a processor configured to: receive configuration information, wherein the configuration information indicates an input data type, a set of reference vectors, and a distance metric type; receive a channel state information (CSI) reference signal (CSI-RS); determine a measured CSI based on measurement values associated with the CSI-RS; generate compressed CSI based on the measured CSI and the input data type; calculate distance metrics associated with the measured CSI and reference vectors of the set of reference vectors based on the distance metric type; and send a report to a network node, wherein the report indicates the compressed CSI and the distance metrics associated with the measured CSI and reference vectors of the set of reference vectors.

2. The WTRU of claim 1, wherein the processor being configured to calculate the distance metrics associated with the measured CSI and reference vectors of the set of reference vectors comprises the processor being configured to calculate respective distance metrics between the measured CSI and each reference vector of the set of reference vectors.

3. The WTRU of claim 1, wherein the distance metric type is a normalized mean square error (NMSE), and the processor being configured to calculate the distance metrics associated with the measured CSI and reference vectors of the set of reference vectors comprises the processor being configured to calculate NMSEs of the measured CSI and reference vectors of the set of reference vectors.

4. The WTRU of claim 1, wherein the distance metric type is a cosine similarity, and the processor being configured to calculate the distance metrics associated with the measured CSI and reference vectors of the set of reference vectors comprises the processor being configured to calculate cosine similarities of the measured CSI and reference vectors of the set of reference vectors.

5. The WTRU of claim 1, wherein the processor being configured to calculate the distance metrics associated with the measured CSI and reference vectors of the set of reference vectors comprises the processor being configured to calculate respective distance metrics between the measured CSI and each reference vector of the set of reference vectors, and wherein the report further indicates a minimum distance metric of the respective distance metrics or a maximum distance metric of the respective distance metrics.

6. The WTRU of claim 1, wherein the distance metric type comprises a function that maps an input tensor to a scalar value, wherein the distance metrics associated with the measured CSI and reference vectors of the set of reference vectors comprise the scalar values.

7. The WTRU of claim 1, wherein the reference vectors are associated with a first domain, the distance metric type is associated with a second domain, and the processor is further configured to transform the reference vectors from the first domain to the second domain.

8. The WTRU of claim 1, wherein the input data type comprises a full channel matrix or an eigenvector.

9. A method for execution by a wireless transmit / receive unit (WTRU), the method comprising: ​ receiving configuration information, wherein the configuration information indicates an input data type, a set of reference vectors, and a distance metric type; receiving channel state information (CSI) reference signals (CSI-RS); determining measured CSI based on measurement values associated with the CSI-RS; generating compressed CSI based on the measured CSI and the input data type; calculating distance metrics associated with the measured CSI and reference vectors of the set of reference vectors based on the distance metric type; and sending a report to a network node, wherein the report indicates the compressed CSI and the distance metrics associated with the measured CSI and reference vectors of the set of reference vectors.

10. The method of claim 9, wherein calculating the distance metrics associated with the measured CSI and reference vectors of the set of reference vectors comprises calculating respective distance metrics between the measured CSI and each reference vector of the set of reference vectors.

11. The method of claim 9, wherein the distance metric type is a normalized mean square error (NMSE), and calculating the distance metrics associated with the measured CSI and reference vectors of the set of reference vectors comprises calculating the NMSE of the measured CSI and the reference vectors of the set of reference vectors.

12. The method of claim 9, wherein the distance metric type is a cosine similarity, and calculating the distance metrics associated with the measured CSI and reference vectors of the set of reference vectors comprises calculating the cosine similarity of the measured CSI and the reference vectors of the set of reference vectors.

13. The method of claim 9, wherein calculating the distance metrics associated with the measured CSI and reference vectors of the set of reference vectors comprises calculating respective distance metrics between the measured CSI and each reference vector of the set of reference vectors, and wherein the report further indicates a minimum distance metric of the respective distance metrics or a maximum distance metric of the respective distance metrics.

14. The method of claim 9, wherein the distance metric type comprises a function that maps an input tensor to a scalar value, wherein the distance metrics associated with the measured CSI and reference vectors of the set of reference vectors comprise the scalar values.

15. The method of claim 9, wherein the reference vectors are associated with a first domain, the distance metric type is associated with a second domain, and the method further comprises transforming the reference vectors from the first domain to the second domain.

16. The method of claim 9, wherein the input data type comprises a full channel matrix or an eigenvector.

Images