EP4681339A1 - Aiml enabled csi feedback with multiple-spatial-stream for wlan systems - Google Patents

Abstract

Methods and devices in which an access point, AP, sends, to one or more stations, information indicative of a request for channel state information, CSI, receives, from at least one station, information indicative of CSI, clusters received CSI to obtain indexed information for a set of clusters, and sends information indicative of the indexed information for the set of clusters to at least one station to be used for CSI feedback procedure.

Description

AIML ENABLED CSI FEEDBACK WITH MULTIPLE-SPATIAL-STREAM FOR WLAN SYSTEMS

CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of U.S. Provisional Application No. 63/490,899, filed 17 March 2023 and of U.S. Provisional Application No. 63/463,465, filed 2 May 2023, which are incorporated herein by reference in their entirety.

BACKGROUND

[0002] Artificial intelligent (Al) and machine learning (ML) have been recently introduced for improving efficiencies and bandwidth capabilities of wireless networks such as a wireless local area network (WLAN). When AIML-enabled channel state (CSI) clustering is performed in an access point (AP) and/or a non-AP station (STA), using different CSI clustering parameters, e.g., clustering criteria, inputs of the clustering, representation of the centroid of each cluster, etc., significant performance differences may arise. It would be desirable to have a set of uniform CSI clustering parameters among the AI/ML enabled STAs, e.g., AP and/or non-AP STAs, and/or APs that are in a same coordinated set. Further, it would be desirable to enable these parameter exchanges between AP and non-AP STAs.

[0003] Additionally, with current CSI feedback algorithms, there is no indication to differentiate the CSI feedback between single user (SU)-multiple input multiple output (MIMO) and multi-user (MU)-MIMO. Furthermore, when a STA is paired with different STAs in MU-MI MO transmission, the channel quality may change significantly. It would be desirable to provide capabilities to enable AIML-based MU-MIMO CSI feedback to provide accurate channel information to an AP, such that the AP may better group STAs in one MU-MIMO transmission.

[0004] In a coordinated multi-AP set, a joint transmission from multi-APs may be performed. To enable the joint transmission from multiple-APs, the participating APs may need to obtain the channel information from unassociated STAs. There is a need for APs to be able to negotiate the AIML parameters used for CSI compression and make an agreement on the AIML parameters.

SUMMARY

[0005] Aspects features and advantages of the disclosed embodiments, may address one or more of the foregoing needs or desires through methods and devices for CSI clustering, AIML based MIMO CSI feedback and AIML Training algorithms described hereinafter.

[0006] In a first aspect, the present principles are directed to a method for an access point, AP, the method comprising sending, to one or more stations, information indicative of a request for channel state information, CSI, receiving, from at least one station, information indicative of CSI, clustering received CSI to obtain indexed information for a set of clusters, and sending information indicative of the indexed information for the set of clusters to at least one station to be used for CSI feedback procedure.

[0007] In a second aspect, the present principles are directed to an access point, AP, comprising at least one processor configured to send, to one or more stations, information indicative of a request for channel state information, CSI, receive, from at least one station, information indicative of CSI, cluster received CSI to obtain indexed information for a set of clusters, and send information indicative of the indexed information for the set of clusters to at least one station to be used for CSI feedback procedure.

[0008] In a third aspect, the present principles are directed to a method for a wireless station associated with an access point, the method comprising receiving, from the access point, information indicative of a request for information about a set of channel state information, CSI, clusters, and sending, to the access point, requested cluster indices.

[0009] In a fourth aspect, the present principles are directed to a wireless station configured for association with an access point, the wireless station comprising at least one processor configured to receive, from the access point, information indicative of a request for information about a set of channel state information, CSI, clusters, and send, to the access point, requested cluster indices.

BRIEF DESCRIPTION OF THE DRAWINGS

[0010] A more detailed understanding may be had from the following description, given by way of example in conjunction with the accompanying drawings, wherein like reference numerals in the figures indicate like elements, and wherein:

[0011] FIG. 1A is a system diagram illustrating an example communications system in which one or more disclosed embodiments may be implemented;

[0012] FIG. 1 B is a system diagram illustrating an example wireless transmit/receive unit (WTRU) that may be used within the communications system illustrated in FIG. 1 A according to an embodiment;

[0013] FIG. 1C is a system diagram illustrating an example radio access network (RAN) and an example core network (CN) that may be used within the communications system illustrated in FIG. 1A according to an embodiment;

[0014] FIG. 1 D is a system diagram illustrating a further example RAN and a further example CN that may be used within the communications system illustrated in FIG. 1 A according to an embodiment;

[0015] FIG. 2 is a message sequence diagram showing sequential versus joint channel sounding in Multi-AP WLAN;

[0016] FIG. 3 shows a high efficiency (HE) null data packet (NDP) Announcement frame format;

[0017] FIG. 4 shows a STA info field format in an extremely high throughput (EHT) NDP Announcement frame;

[0018] FIG. 5 shows an example Trigger Frame format;

[0019] FIG. 6 shows an EHT Variant User Info field format; [0020] FIG. 7 shows an EHT Special User Info field format;

[0021] FIG. 8 is a functional block diagram showing CSI feedback candidate generation using K-means clustering according to one embodiment;

[0022] FIG. 9 is a functional block diagram showing generation of an index of the closest candidate for feedback;

[0023] FIG. 10 is a sequence diagram for Al ML CSI Compression Model Parameter Exchange according to one embodiment;

[0024] FIG. 11 Trigger Dependent User Info subfield format in an Al ML Parameters Report Poll (APRP) according to an embodiment;

[0025] FIG. 12 is an example AIML Control Information subfield according to an example embodiment;

[0026] FIG. 13 is a sequence diagram for an AIML CSI Compression Parameters Report Operation of an embodiment;

[0027] FIG. 14 is a message diagram of an example two-round enhanced EHT sounding operation for MU-MIMO according to an embodiment;

[0028] FIG. 15 is a sequence diagram showing an example of multi-round EHT sounding operation for CQI reports according to an embodiment;

[0029] FIG. 16 is a message diagram of another example two-round enhanced EHT sounding operation for MU- MIMO according to an embodiment;

[0030] FIG. 17 is a diagram showing separate AIML clustering on two different types of angles according to embodiments;

[0031] FIG. 18 is a diagram showing an example of re-clustering within clusters corresponding to high frequency candidates;

[0032] FIG. 19 is a diagram showing an example of selecting new candidate from combined subsets of low frequency data; and

[0033] FIG. 20 illustrates a polar diagram showing effective distance of two angles.

DETAILED DESCRIPTION

[0034] FIG. 1A is a diagram illustrating an example communications system 100 in which one or more disclosed embodiments may be implemented. The communications system 100 may be a multiple access system that provides content, such as voice, data, video, messaging, broadcast, etc., to multiple wireless users. The communications system 100 may enable multiple wireless users to access such content through the sharing of system resources, including wireless bandwidth. For example, the communications systems 100 may employ one or more channel access methods, such as

- o - code division multiple access (CDMA), time division multiple access (TDMA), frequency division multiple access (FDMA), orthogonal FDMA (OFDMA), single-carrier FDMA (SC-FDMA), zero-tail unique-word discrete Fourier transform Spread OFDM (ZT-UW-DFT-S-OFDM), unique word OFDM (UW-OFDM), resource block-filtered OFDM, filter bank multicarrier (FBMC), and the like.

[0035] As shown in FIG. 1A, the communications system 100 may include wireless transmit/receive units (WTRUs) 102a, 102b, 102c, 102d, a radio access network (RAN) 104, a core network (ON) 106, a public switched telephone network (PSTN) 108, the Internet 110, and other networks 112, though it will be appreciated that the disclosed embodiments contemplate any number of WTRUs, base stations, networks, and/or network elements. Each of the WTRUs 102a, 102b, 102c, 102d may be any type of device configured to operate and/or communicate in a wireless environment. By way of example, the WTRUs 102a, 102b, 102c, 102d, any of which may be referred to as a station (STA), may be configured to transmit and/or receive wireless signals and may 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-Fl 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 and/or other wireless devices operating in an industrial and/or an automated processing chain contexts), a consumer electronics device, a device operating on commercial and/or industrial wireless networks, and the like. Any of the WTRUs 102a, 102b, 102c and 102d may be interchangeably referred to as a UE.

[0036] The communications systems 100 may also include a base station 114a and/or a base station 114b. Each of the base stations 114a, 114b may be any type of device configured to wirelessly interface with at least one of the WTRUs 102a, 102b, 102c, 102d to facilitate access to one or more communication networks, such as the CN 106, the Internet 110, and/or the other networks 112. By way of example, the base stations 114a, 114b may be a base transceiver station (BTS), a NodeB, an eNode B (eNB), a Home Node B, a Home eNode B, a next generation NodeB, such as a gNode B (gNB), a new radio (NR) NodeB, a site controller, an access point (AP), a wireless router, 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 may include any number of interconnected base stations and/or network elements.

[0037] The base station 114a may be part of the RAN 104, which may also include other base stations and/or network elements (not shown), such as a base station controller (BSC), a radio network controller (RNC), relay nodes, and the like. The base station 114a and/or the base station 114b may be configured to transmit and/or receive wireless signals on one or more carrier frequencies, which may be referred to as a cell (not shown). These frequencies may be in licensed spectrum, unlicensed spectrum, or a combination of licensed and unlicensed spectrum. A cell may provide coverage for a wireless service to a specific geographical area that may be relatively fixed or that may change over time. The cell may further be divided into cell sectors. For example, the cell associated with the base station 114a may be divided into three sectors Thus, in one embodiment, the base station 114a may include three transceivers, i.e., one for each sector of the cell. In an embodiment, the base station 114a may employ multiple-input multiple output (MIMO) technology and may utilize multiple transceivers for each sector of the cell. For example, beamforming may be used to transmit and/or receive signals in desired spatial directions.

[0038] The base stations 114a, 114b may communicate with one or more of the WTRUs 102a, 102b, 102c, 102d over an air interface 116, which may be any suitable wireless communication link (e.g., radio frequency (RF), microwave, centimeter wave, micrometer wave, infrared (IR), ultraviolet (UV), visible light, etc.). The air interface 116 may be established using any suitable radio access technology (RAT).

[0039] More specifically, as noted above, the communications system 100 may be a multiple access system and may employ one or more channel access schemes, such as CDMA, TDMA, FDMA, OFDMA, SC-FDMA, and the like. For example, the base station 114a in the RAN 104 and the WTRUs 102a, 102b, 102c may implement a radio technology such as Universal Mobile Telecommunications System (UMTS) Terrestrial Radio Access (UTRA), which may establish the air interface 116 using wideband CDMA (WCDMA). WCDMA may include communication protocols such as High-Speed Packet Access (HSPA) and/or Evolved HSPA (HSPA+). HSPA may include High-Speed Downlink (DL) Packet Access (HSDPA) and/or High-Speed Uplink (UL) Packet Access (HSUPA).

[0040] In an embodiment, the base station 114a and the WTRUs 102a, 102b, 102c may implement a radio technology such as Evolved UMTS Terrestrial Radio Access (E-UTRA), which may establish the air interface 116 using Long Term Evolution (LTE) and/or LTE-Advanced (LTE-A) and/or LTE-Advanced Pro (LTE-A Pro).

[0041] In an embodiment, the base station 114a and the WTRUs 102a, 102b, 102c may implement a radio technology such as NR Radio Access , which may establish the air interface 116 using NR.

[0042] In an embodiment, the base station 114a and the WTRUs 102a, 102b, 102c may implement multiple radio access technologies. For example, the base station 114a and the WTRUs 102a, 102b, 102c may implement LTE radio access and NR radio access together, for instance using dual connectivity (DC) principles. Thus, the air interface utilized by WTRUs 102a, 102b, 102c may be characterized by multiple types of radio access technologies and/or transmissions sent to/from multiple types of base stations (e.g., an eNB and a gNB).

[0043] In other embodiments, the base station 114a and the WTRUs 102a, 102b, 102c may implement radio technologies such as IEEE 802.11 (i.e., Wireless Fidelity (WiFi), IEEE 802.16 (i.e., Worldwide Interoperability for Microwave Access (WiMAX)), CDMA2000, CDMA2000 1 X, CDMA2000 EV-DO, Interim Standard 2000 (IS-2000), Interim Standard 95 (I S-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.

[0044] The base station 114b in FIG. 1A may be a wireless router, Home Node B, Home eNode B, or access point, for example, and may utilize any suitable RAT for facilitating wireless connectivity in a localized area, such as a place of business, a home, a vehicle, a campus, an industrial facility, an air corridor (e.g., for use by drones), a roadway, and the like. In one embodiment, the base station 114b and the WTRUs 102c, 102d may implement a radio technology such as IEEE 802.11 to establish a wireless local area network (WLAN). In an embodiment, the base station 114b and the WTRUs 102c, 102d may implement a radio technology such as IEEE 802.15 to establish a wireless personal area network (WPAN). In yet another embodiment, the base station 114b and the WTRUs 102c, 102d may utilize a cellular-based RAT (e.g., WCDMA, CDMA2000, GSM, LTE, LTE-A, LTE-A Pro, NR etc.) to establish a picocell or femtocell. As shown in FIG. 1A, the base station 114b may have a direct connection to the Internet 110. Thus, the base station 114b may not be required to access the Internet 110 via the CN 106.

[0045] The RAN 104 may be in communication with the CN 106, which may be any type of network configured to provide voice, data, applications, and/or voice over internet protocol (VoIP) services to one or more of the WTRUs 102a, 102b, 102c, 102d. The data may have 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 may provide call control, billing services, mobile location-based services, pre-paid calling, Internet connectivity, video distribution, etc., and/or perform high-level security functions, such as user authentication. Although not shown in FIG. 1A, it will be appreciated that the RAN 104 and/or the CN 106 may be in direct or indirect communication with other RANs that employ the same RAT as the RAN 104 or a different RAT. For example, in addition to being connected to the RAN 104, which may be utilizing a NR radio technology, the CN 106 may also be in communication with another RAN (not shown) employing a GSM, UMTS, CDMA 2000, WiMAX, E-UTRA, or WiFi radio technology.

[0046] The CN 106 may also serve as a gateway for the WTRUs 102a, 102b, 102c, 102d to access the PSTN 108, the Internet 110, and/or the other networks 112. The PSTN 108 may include circuit-switched telephone networks that provide plain old telephone service (POTS). The Internet 110 may include a global system of interconnected computer networks and devices that use common communication protocols, such as the transmission control protocol (TCP), user datagram protocol (UDP) and/or the internet protocol (IP) in the TCP/IP internet protocol suite. The networks 112 may include wired and/or wireless communications networks owned and/or operated by other service providers. For example, the networks 112 may include another CN connected to one or more RANs, which may employ the same RAT as the RAN 104 or a different RAT.

[0047] Some or all of the WTRUs 102a, 102b, 102c, 102d in the communications system 100 may include multi-mode capabilities (e.g., the WTRUs 102a, 102b, 102c, 102d may include multiple transceivers for communicating with different wireless networks over different wireless links). For example, the WTRU 102c shown in FIG. 1A may be configured to communicate with the base station 114a, which may employ a cellular-based radio technology, and with the base station 114b, which may employ an IEEE 802 radio technology.

[0048] FIG. 1 B is a system diagram illustrating an example WTRU 102. As shown in FIG. 1 B, the WTRU 102 may 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 source 134, a global positioning system (GPS) chipset 136, and/or other peripherals 138, among others. It will be appreciated that the WTRU 102 may include any sub-combination of the foregoing elements while remaining consistent with an embodiment.

[0049] The processor 118 may be a general purpose processor, a special purpose processor, a conventional processor, a digital signal processor (DSP), a plurality of microprocessors, one or more microprocessors in association with a DSP core, a controller, a microcontroller, Application Specific Integrated Circuits (ASICs), Field Programmable Gate Arrays (FPGAs), any other type of integrated circuit (IC), a state machine, and the like. The processor 118 may perform signal coding, data processing, power control, input/output processing, and/or any other functionality that enables the WTRU 102 to operate in a wireless environment. The processor 118 may be coupled to the transceiver 120, which may be coupled to the transmit/receive element 122. While FIG. 1 B depicts the processor 118 and the transceiver 120 as separate components, it will be appreciated that the processor 118 and the transceiver 120 may be integrated together in an electronic package or chip.

[0050] The transmit/receive element 122 may 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 may be an antenna configured to transmit and/or receive RF signals. In an embodiment, the transmit/receive element 122 may be an emitter/detector configured to transmit and/or receive IR, UV, or visible light signals, for example. In yet another embodiment, the transmit/receive element 122 may be configured to transmit and/or receive both RF and light signals. It will be appreciated that the transmit/receive element 122 may be configured to transmit and/or receive any combination of wireless signals.

[0051] Although the transmit/receive element 122 is depicted in FIG. 1 B as a single element, the WTRU 102 may include any number of transmit/receive elements 122. More specifically, the WTRU 102 may employ MIMO technology. Thus, in one embodiment, the WTRU 102 may include two or more transmit/receive elements 122 (e.g., multiple antennas) for transmitting and receiving wireless signals over the air interface 116.

[0052] The transceiver 120 may be configured to modulate the signals that are to be transmitted by the transmit/receive element 122 and to demodulate the signals that are received by the transmit/receive element 122. As noted above, the WTRU 102 may have multi-mode capabilities. Thus, the transceiver 120 may include multiple transceivers for enabling the WTRU 102 to communicate via multiple RATs, such as NR and IEEE 802.11 , for example.

[0053] The processor 118 of the WTRU 102 may be coupled to, and may 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 may also output user data to the speaker/microphone 124, the keypad 126, and/or the display/touchpad 128. In addition, the processor 118 may access information from, and store data in, any type of suitable memory, such as the non-removable memory 130 and/or the removable memory 132. The non-removable memory 130 may include random-access memory (RAM), read-only memory (ROM), a hard disk, or any other type of memory storage device. The removable memory 132 may include a subscriber identity module (SIM) card, a memory stick, a secure digital (SD) memory card, and the like. In other embodiments, the processor 118 may access information from, and store data in, memory that is not physically located on the WTRU 102, such as on a server or a home computer (not shown).

[0054] The processor 118 may receive power from the power source 134, and may be configured to distribute and/or control the power to the other components in the WTRU 102. The power source 134 may be any suitable device for powering the WTRU 102. For example, the power source 134 may include one or more dry cell batteries (e.g., nickelcadmium (NiCd), nickel-zinc (NiZn), nickel metal hydride (NiMH), lithium-ion (Li-ion), etc.), solar cells, fuel cells, and the like.

[0055] The processor 118 may also be coupled to the GPS chipset 136, which may be configured to provide location information (e.g., longitude and latitude) regarding the current location of the WTRU 102. In addition to, or in lieu of, the information from the GPS chipset 136, the WTRU 102 may receive location information over the air interface 116 from a base station (e.g., base stations 114a, 114b) and/or determine its location based on the timing of the signals being received from two or more nearby base stations. It will be appreciated that the WTRU 102 may acquire location information by way of any suitable location-determination method while remaining consistent with an embodiment.

[0056] The processor 118 may further be coupled to other peripherals 138, which may include one or more software and/or hardware modules that provide additional features, functionality and/or wired or wireless connectivity. For example, the peripherals 138 may include an accelerometer, an e-compass, a satellite transceiver, a digital camera (for photographs and/or video), a universal serial bus (USB) port, a vibration device, a television transceiver, a hands free headset, a Bluetooth® module, a frequency modulated (FM) radio unit, a digital music player, a media player, a video game player module, an Internet browser, a Virtual Reality and/or Augmented Reality (VR/AR) device, an activity tracker, and the like. The peripherals 138 may include one or more sensors. The sensors may be one or more of a gyroscope, an accelerometer, a hall effect sensor, a magnetometer, an orientation sensor, a proximity sensor, a temperature sensor, a time sensor; a geolocation sensor, an altimeter, a light sensor, a touch sensor, a magnetometer, a barometer, a gesture sensor, a biometric sensor, a humidity sensor and the like.

[0057] The WTRU 102 may include a full duplex radio for which transmission and reception of some or all of the signals (e.g., associated with particular subframes for both the UL (e.g., for transmission) and DL (e.g., for reception) may be concurrent and/or simultaneous. The full duplex radio may include an interference management unit to reduce and or substantially eliminate self-interference via either hardware (e.g., a choke) or signal processing via a processor (e.g., a separate processor (not shown) or via processor 118). In an embodiment, the WTRU 102 may include a half-duplex radio for which transmission and reception of some or all of the signals (e.g., associated with particular subframes for either the UL (e.g., for transmission) or the DL (e.g., for reception)).

[0058] FIG. 1C is a system diagram illustrating the RAN 104 and the CN 106 according to an embodiment. As noted above, the RAN 104 may employ an E-UTRA radio technology to communicate with the WTRUs 102a, 102b, 102c over the air interface 116. The RAN 104 may also be in communication with the CN 106. [0059] The RAN 104 may include eNode-Bs 160a, 160b, 160c, though it will be appreciated that the RAN 104 may include any number of eNode-Bs while remaining consistent with an embodiment. The eNode-Bs 160a, 160b, 160c may each include one or more transceivers for communicating with the WTRUs 102a, 102b, 102c over the air interface 116. In one embodiment, the eNode-Bs 160a, 160b, 160c may implement Ml MO technology. Thus, the eNode-B 160a, for example, may use multiple antennas to transmit wireless signals to, and/or receive wireless signals from, the WTRU 102a.

[0060] Each of the eNode-Bs 160a, 160b, 160c may be associated with a particular cell (not shown) and may be configured to handle radio resource management decisions, handover decisions, scheduling of users in the UL and/or DL, and the like. As shown in FIG. 1C, the eNode-Bs 160a, 160b, 160c may communicate with one another over an X2 interface. [0061] The CN 106 shown in FIG. 1C may include a mobility management entity (MME) 162, a serving gateway (SGW) 164, and a packet data network (PDN) gateway (PGW) 166. While the foregoing elements are depicted as part of the CN 106, it will be appreciated that any of these elements may be owned and/or operated by an entity other than the CN operator.

[0062] The MME 162 may be connected to each of the eNode-Bs 162a, 162b, 162c in the RAN 104 via an S1 interface and may serve as a control node. For example, the MME 162 may be responsible for authenticating users of the WTRUs 102a, 102b, 102c, bearer activation/deactivation, selecting a particular serving gateway during an initial attach of the WTRUs 102a, 102b, 102c, and the like. The MME 162 may provide a control plane function for switching between the RAN 104 and other RANs (not shown) that employ other radio technologies, such as GSM and/or WCDMA.

[0063] The SGW 164 may be connected to each of the eNode Bs 160a, 160b, 160c in the RAN 104 via the S1 interface. The SGW 164 may generally route and forward user data packets to/from the WTRUs 102a, 102b, 102c. The SGW 164 may perform other functions, such as anchoring 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.

[0064] The SGW 164 may be connected to the PGW 166, which may provide the WTRUs 102a, 102b, 102c with access to packet-switched networks, such as the Internet 110, to facilitate communications between the WTRUs 102a, 102b, 102c and IP-enabled devices.

[0065] The CN 106 may facilitate communications with other networks. For example, the CN 106 may provide the WTRUs 102a, 102b, 102c with access to circuit-switched networks, such as the PSTN 108, to facilitate communications between the WTRUs 102a, 102b, 102c and traditional land-line communications devices. For example, the CN 106 may include, or may communicate with, an IP gateway (e.g., an IP multimedia subsystem (IMS) server) that serves as an interface between the CN 106 and the PSTN 108. In addition, the CN 106 may provide the WTRUs 102a, 102b, 102c with access to the other networks 112, which may include other wired and/or wireless networks that are owned and/or operated by other service providers. [0066] Although the WTRU is described in FIGS. 1A-1 D as a wireless terminal, it is contemplated that in certain representative embodiments that such a terminal may use (e.g., temporarily or permanently) wired communication interfaces with the communication network.

[0067] In representative embodiments, the other network 112 may be a WLAN.

[0068] A WLAN in Infrastructure Basic Service Set (BSS) mode may have an Access Point (AP) for the BSS and one or more stations (STAs) associated with the AP. The AP may have access or an interface to a Distribution System (DS) or another type of wired/wireless network that carries traffic in to and/or out of the BSS. Traffic to STAs that originates from outside the BSS may arrive through the AP and may be delivered to the STAs. Traffic originating from STAs to destinations outside the BSS may be sent to the AP to be delivered to respective destinations. Traffic between STAs within the BSS may be sent through the AP, for example, where the source STA may send traffic to the AP and the AP may deliver the traffic to the destination STA. The traffic between STAs within a BSS may be considered and/or referred to as peer-to-peer traffic. The peer-to-peer traffic may be sent between (e.g., directly between) the source and destination STAs with a direct link setup (DLS). In certain representative embodiments, the DLS may use an 802.11e DLS or an 802.11z tunneled DLS (TDLS). A WLAN using an Independent BSS (IBSS) mode may not have an AP, and the STAs (e.g., all of the STAs) within or using the IBSS may communicate directly with each other. The IBSS mode of communication may sometimes be referred to herein as an “ad-hoc” mode of communication.

[0069] When using the 802.11 ac infrastructure mode of operation or a similar mode of operations, the AP may transmit a beacon on a fixed channel, such as a primary channel. The primary channel may be a fixed width (e.g., 20 MHz wide bandwidth) or a dynamically set width. The primary channel may be the operating channel of the BSS and may be used by the STAs to establish a connection with the AP. In certain representative embodiments, Carrier Sense Multiple Access with Collision Avoidance (CSMA/CA) may be implemented, for example in 802.11 systems. For CSMA/CA, the STAs (e.g., every STA), including the AP, may sense the primary channel. If the primary channel is sensed/detected and/or determined to be busy by a particular STA, the particular STA may back off. One STA (e.g., only one station) may transmit at any given time in a given BSS.

[0070] High Throughput (HT) STAs may use a 40 MHz wide channel 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.

[0071] Very High Throughput (VHT) STAs may support 20MHz, 40 MHz, 80 MHz, and/or 160 MHz wide channels. The 40 MHz, and/or 80 MHz, channels may be formed by combining contiguous 20 MHz channels A 160 MHz channel may be formed by combining 8 contiguous 20 MHz channels, or by combining two non-contiguous 80 MHz channels, which may be referred to as an 80+80 configuration. For the 80+80 configuration, the data, after channel encoding, may be passed through a segment parser that may divide the data into two streams. Inverse Fast Fourier Transform (IFFT) processing, and time domain processing, may be done on each stream separately. The streams may be mapped on to the two 80 MHz channels, and the data may be transmitted by a transmitting STA. At the receiver of the receiving STA, the above described operation for the 80+80 configuration may be reversed, and the combined data may be sent to the Medium Access Control (MAC).

[0072] Sub 1 GHz modes of operation are supported by 802.11 af and 802.11 ah. The channel operating bandwidths, and carriers, are reduced in 802.11 af and 802.11 ah relative to those used in 802.11 n, and 802.11ac. 802.11af supports 5 MHz, 10 MHz, and 20 MHz bandwidths in the TV White Space (TVWS) spectrum, and 802.11 ah supports 1 MHz, 2 MHz, 4 MHz, 8 MHz, and 16 MHz bandwidths using non-TVWS spectrum. According to a representative embodiment, 802.11 ah may support Meter Type Control/Machine-Type Communications (MTC), such as MTC devices in a macro coverage area. MTC devices may have certain capabilities, for example, limited capabilities including support for (e.g., only support for) certain and/or limited bandwidths. The MTC devices may include a battery with a battery life above a threshold (e.g., to maintain a very long battery life).

[0073] WLAN systems, which may support multiple channels, and channel bandwidths, such as 802.11 n, 802.11ac, 802.11 af, and 802.11 ah, include a channel which may be designated as the primary channel. The primary channel may have a bandwidth equal to the largest common operating bandwidth supported by all STAs in the BSS. The bandwidth of the primary channel may be set and/or limited by a STA, from among all STAs in operating in a BSS, which supports the smallest bandwidth operating mode. In the example of 802.11 ah, the primary channel may be 1 MHz wide for STAs (e.g., MTC type devices) that support (e.g., only support) a 1 MHz mode, even if the AP, and other STAs in the BSS support 2 MHz, 4 MHz, 8 MHz, 16 MHz, and/or other channel bandwidth operating modes. Carrier sensing and/or Network Allocation Vector (NAV) settings may depend on the status of the primary channel. If the primary channel is busy, for example, due to a STA (which supports only a 1 MHz operating mode) transmitting to the AP, all available frequency bands may be considered busy even though a majority of the available frequency bands remains idle.

[0074] In the United States, the available frequency bands, which may be used by 802.11 ah, are from 902 MHz to 928 MHz. In Korea, the available frequency bands are from 917.5 MHz to 923 5 MHz. In Japan, the available frequency bands are from 916.5 MHz to 927.5 MHz. The total bandwidth available for 802.11 ah is 6 MHz to 26 MHz depending on the country code.

[0075] FIG. 1 D is a system diagram illustrating the RAN 104 and the CN 106 according to an embodiment. As noted above, the RAN 104 may employ an NR radio technology to communicate with the WTRUs 102a, 102b, 102c over the air interface 116. The RAN 104 may also be in communication with the CN 106.

[0076] The RAN 104 may include gNBs 180a, 180b, 180c, though it will be appreciated that the RAN 104 may include any number of gNBs while remaining consistent with an embodiment. The gNBs 180a, 180b, 180c may each include one or more transceivers for communicating with the WTRUs 102a, 102b, 102c over the air interface 116. In one embodiment, the gNBs 180a, 180b, 180c may implement Ml MO technology. For example, gNBs 180a, 108b may utilize beamforming to transmit signals to and/or receive signals from the gNBs 180a, 180b, 180c. Thus, the gNB 180a, for example, may use multiple antennas to transmit wireless signals to, and/or receive wireless signals from, the WTRU 102a. In an embodiment, the gNBs 180a, 180b, 180c may implement carrier aggregation technology. For example, the gNB 180a may transmit multiple component carriers to the WTRU 102a (not shown). A subset of these component carriers may be on unlicensed spectrum while the remaining component carriers may be on licensed spectrum. In an embodiment, the gNBs 180a, 180b, 180c may implement Coordinated Multi-Point (CoMP) technology. For example, WTRU 102a may receive coordinated transmissions from gNB 180a and gNB 180b (and/or gNB 180c).

[0077] The WTRUs 102a, 102b, 102c may communicate with gNBs 180a, 180b, 180c using transmissions associated with a scalable numerology. For example, the OFDM symbol spacing and/or OFDM subcarrier spacing may vary for different transmissions, different cells, and/or different portions of the wireless transmission spectrum. The WTRUs 102a, 102b, 102c may communicate with gNBs 180a, 180b, 180c using subframe or transmission time intervals (TTIs) of various or scalable lengths (e.g., containing a varying number of OFDM symbols and/or lasting varying lengths of absolute time).

[0078] The gNBs 180a, 180b, 180c may be configured to communicate with the WTRUs 102a, 102b, 102c in a standalone configuration and/or a non-standalone configuration. In the standalone configuration, WTRUs 102a, 102b, 102c may communicate with gNBs 180a, 180b, 180c without also accessing other RANs (e.g., such as eNode-Bs 160a, 160b, 160c). In the standalone configuration, WTRUs 102a, 102b, 102c may utilize one or more of gNBs 180a, 180b, 180c as a mobility anchor point. In the standalone configuration, WTRUs 102a, 102b, 102c may communicate with gNBs 180a, 180b, 180c using signals in an unlicensed band. In a non-standalone configuration WTRUs 102a, 102b, 102c may communicate with/connect to gNBs 180a, 180b, 180c while also communicating with/connecting to another RAN such as eNode-Bs 160a, 160b, 160c. For example, WTRUs 102a, 102b, 102c may implement DC principles to communicate with one or more gNBs 180a, 180b, 180c and one or more eNode-Bs 160a, 160b, 160c substantially simultaneously. In the non-standalone configuration, eNode-Bs 160a, 160b, 160c may serve as a mobility anchor for WTRUs 102a, 102b, 102c and gNBs 180a, 180b, 180c may provide additional coverage and/or throughput for servicing WTRUs 102a, 102b, 102c.

[0079] Each of the gNBs 180a, 180b, 180c may be associated with a particular cell (not shown) and may be configured to handle radio resource management decisions, handover decisions, scheduling of users in the UL and/or DL, support of network slicing, DC, interworking between NR and E-UTRA, routing of user plane data towards User Plane Function (UPF) 184a, 184b, routing of control plane information towards Access and Mobility Management Function (AMF) 182a, 182b and the like. As shown in FIG. 1D, the gNBs 180a, 180b, 180c may communicate with one another over an Xn interface.

[0080] The CN 106 shown in FIG. 1 D may include at least one AMF 182a, 182b, at least one UPF 184a, 184b, at least one Session Management Function (SMF) 183a, 183b, and possibly a Data Network (DN) 185a, 185b. While the foregoing elements are depicted as part of the CN 106, it will be appreciated that any of these elements may be owned and/or operated by an entity other than the CN operator.

[0081] The AMF 182a, 182b may be connected to one or more of the gNBs 180a, 180b, 180c in the RAN 104 via an N2 interface and may serve as a control node. For example, the AMF 182a, 182b may be responsible for authenticating users of the WTRUs 102a, 102b, 102c, support for network slicing (e.g., handling of different protocol data unit (PDU) sessions with different requirements), selecting a particular SMF 183a, 183b, management of the registration area, termination of non-access stratum (NAS) signaling, mobility management, and the like. Network slicing may be used by the AM F 182a, 182b in order to customize CN support for WTRUs 102a, 102b, 102c based on the types of services being utilized WTRUs 102a, 102b, 102c. For example, different network slices may be established for different use cases such as services relying on ultra-reliable low latency (URLLC) access, services relying on enhanced massive mobile broadband (eMBB) access, services for MTC access, and the like. The AMF 182a, 182b may provide a control plane function for switching between the RAN 104 and other RANs (not shown) that employ other radio technologies, such as LTE, LTE-A, LTE-A Pro, and/or non-3GPP access technologies such as WiFi.

[0082] The SMF 183a, 183b may be connected to an AMF 182a, 182b in the CN 106 via an N11 interface. The SMF 183a, 183b may also be connected to a UPF 184a, 184b in the CN 106 via an N4 interface. The SMF 183a, 183b may select and control the UPF 184a, 184b and configure the routing of traffic through the UPF 184a, 184b. The SMF 183a, 183b may perform other functions, such as managing and allocating UE IP address, managing PDU sessions, controlling policy enforcement and QoS, providing DL data notifications, and the like. A PDU session type may be IP-based, non-IP based, Ethernet-based, and the like.

[0083] The UPF 184a, 184b may be connected to one or more of the gNBs 180a, 180b, 180c in the RAN 104 via an N3 interface, which may provide the WTRUs 102a, 102b, 102c with access to packet-switched networks, such as the Internet 110, to facilitate communications between the WTRUs 102a, 102b, 102c and IP-enabled devices. The UPF 184, 184b may perform other functions, such as routing and forwarding packets, enforcing user plane policies, supporting multihomed PDU sessions, handling user plane QoS, buffering DL packets, providing mobility anchoring, and the like.

[0084] The CN 106 may facilitate communications with other networks. For example, the CN 106 may include, or may communicate with, an IP gateway (e.g. , an IP multimedia subsystem (IMS) server) that serves as an interface between the CN 106 and the PSTN 108. In addition, the CN 106 may provide the WTRUs 102a, 102b, 102c with access to the other networks 112, which may 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 may be connected to a local DN 185a, 185b through the UPF 184a, 184b via the N3 interface to the UPF 184a, 184b and an N6 interface between the UPF 184a, 184b and the DN 185a, 185b.

[0085] In view of FIGs. 1A-1 D, and the corresponding description of FIGs. 1A-1 D, one or more, or all, of the functions described herein with regard to one or more of: WTRU 102a-d, Base Station 114a-b, eNode-B 160a-c, MME 162, SGW 164, PGW 166, gNB 180a-c, AMF 182a-b, UPF 184a-b, SMF 183a-b, DN 185a-b, and/or any other device(s) described herein, may be performed by one or more emulation devices (not shown). The emulation devices may be one or more devices configured to emulate one or more, or all, of the functions described herein. For example, the emulation devices may be used to test other devices and/or to simulate network and/or WTRU functions.

[0086] The emulation devices may be designed to implement one or more tests of other devices in a lab environment and/or in an operator network environment. For example, the one or more emulation devices may perform the one or more, or all, functions while being fully or partially implemented and/or deployed as part of a wired and/or wireless communication network in order to test other devices within the communication network. The one or more emulation devices may perform the one or more, or all, functions while being temporarily implemented/deployed as part of a wired and/or wireless communication network. The emulation device may be directly coupled to another device for purposes of testing and/or performing testing using over-the-air wireless communications.

[0087] The one or more emulation devices may perform the one or more, including all, functions while not being implemented/deployed as part of a wired and/or wireless communication network. For example, the emulation devices may be utilized in a testing scenario in a testing laboratory and/or a non-deployed (e.g., testing) wired and/or wireless communication network in order to implement testing of one or more components. The one or more emulation devices may be test equipment. Direct RF coupling and/or wireless communications via RF circuitry (e.g., which may include one or more antennas) may be used by the emulation devices to transmit and/or receive data.

[0088] A WLAN in Infrastructure Basic Service Set (BSS) mode has an Access Point (AP) for the BSS and one or more stations (STAs) associated with the AP. The AP typically has access or interface to a Distribution System (DS) or another type of wired/wireless network that carries traffic in and out of the BSS. T raffic to STAs that originates from outside the BSS arrives through the AP and is delivered to the STAs. Traffic originating from STAs to destinations outside the BSS is sent to the AP to be delivered to the respective destinations. Traffic between STAs within the BSS may also be sent through the AP where the source STA sends traffic to the AP and the AP delivers the traffic to the destination STA.

[0089] Using the 802 11 ac infrastructure mode of operation, the AP may transmit a beacon on a fixed channel, usually the primary channel. This channel may be 20MHz wide, and is the operating channel of the BSS. This channel is also used by the STAs to establish a connection with the AP. The fundamental channel access mechanism in an 802.11 system is Carrier Sense Multiple Access with Collision Avoidance (CSMA/CA). In this mode of operation, every STA, including the AP, will sense the primary channel. If the channel is detected to be busy, the STA backs off. Hence only one STA may transmit at any given time in a given BSS.

[0090] In 802.11n, High Throughput (HT) STAs may also use a 40MHz wide channel for communication. This is achieved by combining the primary 20MHz channel, with an adjacent 20MHz channel to form a 40MHz wide contiguous channel.

[0091] In 802.11ac, Very High Throughput (VHT) STAs may support 20MHz, 40MHz, 80MHz, and 160MHz wide channels. The 40MHz, and 80MHz, channels are formed by combining contiguous 20MHz channels similar to 802.11 n described above. A160MHz channel may be formed either by combining eight contiguous 20MHz channels, or by combining two non-contiguous 80MHz channels, which may also be referred to as an 80+80 configuration. For the 80+80 configuration, the data, after channel encoding, is passed through a segment parser that divides it into two streams. The Inverse Discrete Fourier Transformation (IDFT) operation and time domain processing are done on each stream separately. The streams are then mapped on to the two channels, and the data is transmitted. At the receiver, this mechanism is reversed, and the combined data is sent to the MAC.

[0092] To improve spectral efficiency 802.11 ac has introduced the concept for downlink Multi-User MIMO (MU-MIMO) transmission to multiple STA’s in the same symbol’s time frame, e.g. during a downlink OFDM symbol. The potential for the use of downlink MU-MIMO is also currently considered for 802.11 ah. It is important to note that since downlink MU- MI MO, as it is used in 802.11 ac, uses the same symbol timing to multiple STA's interference of the waveform transmissions to multiple STA’s is not an issue. However, all STA’s involved in MU-MIMO transmission with the AP must use the same channel or band, this limits the operating bandwidth to the smallest channel bandwidth that is supported by the STA’s which are included in the MU-MIMO transmission with the AP.

[0093] IEEE 802.11 be is referred to as Extremely High Throughput (EHT). EHT further increases peak throughput and improves efficiency of the IEEE 802.11 networks. EHT addresses primary use cases and applications including high throughput and low latency applications such as Video-over-WLAN, Augmented Reality (AR) and Virtual Reality (VR).

[0094] Features related to 802.11be EHT include: Multi-AP, Multi-Band/multi-link, 320MHz bandwidth, 16-Spatial Streams, hybrid automatic repeat request (HARQ), AP Coordination and new designs for 6-GHz channel access, among others.

[0095] EHT STAs use the EHT sounding protocol to determine the channel state information. The EHT sounding protocol provides explicit feedback mechanisms, defined as EHT non-trigger-based (non-TB) sounding and EHT triggerbased (TB) sounding, where the EHT beamformer (BFer) determines the channel state by transmitting a training signal (i.e. , an EHT sounding NDP) to the EHT beamformee (BFee), which sends back a transformed estimate of the channel state. The EHT beamformer uses this estimate to derive the steering matrix.

[0096] The EHT beamformee returns an estimate of the channel state in an EHT compressed beamforming/CQI report carried in one or more EHT Compressed Beamforming/CQI frames. There are three types of EHT compressed beamforming/CQI reports. For SU feedback, the EHT compressed beamforming/CQI report consists of an EHT Compressed Beamforming Report field. For MU feedback, the EHT compressed beamforming/CQI report consists of an EHT Compressed Beamforming Report field and EHT MU Exclusive Beamforming Report field. For CQI feedback, the EHT compressed beamforming/CQI report consists of an EHT CQI Report field.

[0097] 802.11 be Multi-AP Transmission defines schemes for Coordinated multi-AP (C-MAP) transmissions including: Coordinated Multi-AP OFDMA; Coordinated Multi-AP TDMA; Coordinated Multi-AP Spatial Reuse; Coordinated beamforming/nulling; and Joint Transmission. In the context of coordinated Multi-AP, several terminologies have been defined including: Sharing AP-an EHT AP which obtains a Transmission Opportunity (TXOP) and initiates the multi-AP coordination; Shared AP-an EHT AP which is coordinated for the multi-AP transmission by the sharing AP; and AP candidate set- a set of APs that may initiate or participate in multi-AP coordination

[0098] In 11 be, mechanisms to determine whether an AP is part of an AP candidate set and can participate as a shared AP in coordinated AP transmission initiated by a sharing AP are still being defined. Further, how an AP shares its frequency/time resources of an obtained TXOP with a set of APs is also being defined. An AP that intends to use the resource (i.e., frequency or time) shared by another AP will be able to indicate its resource needs to the AP that shared the resource. Coordinated OFDMA is supported in 11be, and in a coordinated OFDMA, both DL OFDMA and its corresponding UL OFDMA acknowledgement are allowed. [0099] Channel sounding in 802.11 n and 802.11 ac is performed using two different schemes, explicit or implicit. In explicit channel sounding, the AP transmits an NDP to the STA with a preamble that allows the STA to measure its own channel and send CSI feedback to the AP. In implicit channel sounding, the STA sends an NDP, and the AP measures the channel of the STA assuming that the channel is reciprocal.

[0100] 802.11 be supports a maximum of 16-spatial streams for SU-MIMO and for MU-MI MO, the maximum number of spatial streams allocated to each MU-MI MO scheduled non-AP STA is limited to four. The maximum number of users spatially multiplexed for DL transmissions is eight per resource unit (RU).

[0101] 802.11 be supports two modes of channel sounding in Multiple-AP, sequential sounding, and joint sounding. In sequential sounding, each AP transmits a Null Data Packet (NDP) independently without overlapped sounding period of each AP. Also, joint sounding is provided as optional mode for Multiple-AP, where when an AP has less or equal to a total of eight antennas, all antennas active on all long training field (LTF) tones and uses 802.11 ax P-matrix across OFDM symbols. 802.11 ax is referred to as high efficiency (HE) WLAN.

[0102] The CSI feedback collection can be performed using 802.11ax-like 4-step sounding sequence (NDP announcement (NDPA) + NDP + beamforming report (BFRP) trigger frame (TF) + CSI report) in Multiple-AP to collect the feedback from both in-BSS and overlapping BSS (OBSS) STAs. Further, in sequential sounding for Multiple-AP, a STA can process an NDPA frame and the BFRP Trigger frame received from the OBSS AP and the STA can respond with the corresponding CSI to the OBSS AP, if polled by the BFRP TF from the OBSS AP.

[0103] In sequential sounding, each AP in the coordinating group transmits an NDP in a different non-overlapped time to all the STAs in the coordinating group (i.e. time-multiplexed). In joint sounding, the coordinated APs can transmit the NDP simultaneously where different LTF tones are either spanning the entire bandwidth and multiplexed spatially or using orthogonal codes or otherwise, and the LTF tones are only sent on selected tones for each AP. The two options of Multiple- AP channel sounding are illustrated in FIG. 2.

[0104] When a STA receives an NDP, it measures the channel and prepares the CSI feedback report. Three different ways are proposed to collect the CSI from the STAs: (i) Each AP collects all CSI which includes the feedback of the in-BSS and OBSS stations; (ii) Each AP collects CSI from its associated STAs only; or (iii) The Sharing AP collects the CSI for all the Shared APs in the coordination group.

[0105] There are several challenges of Channel Sounding in Multi-AP including: STAs involved in the sounding cannot hear the coordinating AP (or the master) AP; synchronization of APs in the Multi-AP coordinating set is difficult; overhead, complexity and performance of different sounding schemes; handling variants of NDP Transmission in explicit and implicit sounding; and/or feedback collection and reduction.

[0106] The 802.11 be NDP Announcement (NDPA) is similar to the NDPA of 802.11 ax illustrated in FIG. 3. However, the STA Info field depicted in FIG. 4 is changed to accommodate the new features of EHT. [0107] 802.11 be introduces an Enhanced Trigger Frame. A Trigger Frame was introduced in 802.11ax HE to allocate resources and trigger single or multi-user access in the uplink. The Trigger frame format is shown in FIG. 5. In 802.11 be, a new variant of the User Info field is included, and a Special User Info field is added just after the Common Info field. Both of these enhancements are illustrated in FIG. 6 and FIG. 7 and allow a unified triggering scheme for both HE and EHT devices.

[0108] When Al ML enabled CSI clustering is performed in an AP and/or non-AP STA, using different CSI clustering parameters, e.g., clustering criteria, inputs of the clustering, representation of the centroid of each cluster, significant performance differences may result. It is desirable to have a set of uniform CSI clustering parameters among the AI/ML enabled STAs, e.g., AP and/or non-AP STAs or the APs that are in the same coordinated set, as well as enable these parameter to be exchanged between AP and non-AP STAs.

[0109] Additionally, in the CSI current feedback algorithms, there is no indication to differentiate the CSI feedback between SU-MIMO and MU-MIMO. Furthermore, when a STA is paired with different STAs in MU-MIMO transmission, the channel quality may change significantly. Therefore, the Al ML based MU-MIMO CSI feedback should be enabled to provide accurate channel information to an AP such that the AP may better group STAs in one MU-MIMO transmission.

[0110] Lastly, as mentioned above, in a coordinated multi-AP set, a joint transmission from multi-APs may be performed. To enable the joint transmission from multiple-APs, the participating APs may need to obtain the channel information from unassociated STAs and there is a need for APs to negotiate the AIML parameters used for CSI compression and make an agreement on the AIML parameters.

[0111] Various embodiments relate to methods of AIML Enabled CSI Clustering. In one embodiment, in order for all AIML enabled devices to have a uniform AIML algorithm for CSI compression, an AP may need to indicate the following parameters to other STAs, e.g., non-AP STAs:

[0112] 1. AIML CSI compression method, e.g., clustering method. The AP may specify what part(s) of CSI feedback use compression. For example, the wideband/subband CSI feedback may be used with an index-based method, or the subcarrier based CSI feedback may be used with an index-based method.

[0113] 2. Representation of clustering centroid or candidate (codebook) form: the beamformer (AP) may indicate to the beamformees (non-AP STAs) the form in which the AI/ML based CSI feedback should be represented. As an example, the beamformer may request that the CSI feedback be represented in the form of quantized angle index values, as it is represented in the current standard specification for compressed beamforming. In another example, the CSI feedback may be represented in the form of covariance matrix of the estimated channel (H) generated by the beamformee using the training fields in the NDP frame. The covariance matrix (W) of the given channel estimate may be defined using the K = H^H x H , where (. )^w operation represents the Hermitian of a given matrix. The matrix K may further be normalized by a factor of p-norm of a matrix, i.e., ||W||_p, (where p is the normalization type) which may be implementation dependent (e.g. L1-norm where p=1, L2-norm where p=2, etc.) and may also be defined for a specific implementation depending on the nature of K matrix in cases where multiple spatial streams are being used to transmit data.

[0114] The form of representing the AI/ML based CSI feedback candidates indicated above may also depend on how these CSI feedback candidates are generated. As an example, FIG. 8 shows and example process for CSI feedback candidate generation using the K-means clustering algorithm. In this method, a database 802 contains CSI feedback entries which are represented in the desired form (i.e. quantized angle index vectors, covariance matrices, beamforming feedback matrix, V, etc.). This database may then be fed to the K-means clustering algorithm 804. The K-means clustering method enables the choice of the number of clusters (Af_m) to cluster the data in, and the metric used to calculate the distance between the cluster centroids and each entry in the database. The indicator D_£ may signify this metric (e.g., squared Euclidean distance, Hamming distance, General Cosine Similarity (GCS), etc.). After the K-means algorithm converges, it provides N_m cluster centroids, which may be considered as the CSI feedback candidates. Note the representation form of the cluster centroid and the form of the candidate vector may be same or may not be same. The AP may need to indicate these two forms to the non-AP STA or coordinated AP(s).

[0115] Referring to FIG. 9, in one embodiment, using number of clusters, N_m, after a beamformee (non-AP STA) 904 receives the NDP frame 901 from a beamformer 902, it may compute 903 the steering matrix (V) or the channel matrix (77) depending on how the N_m CSI feedback candidates 913 are represented. If the candidates are represented using covariance matrices, the beamformee may compute the covariance matrix (K) from H. Then the beamformee 904 may perform candidate selection 905 to choose the closest candidate 909 to represent the current V or K. To find the closest candidate, the beamformee may use the same distance metric indicator D_£ 907 as used in the candidate generation. The beamformee may then feed back the index 911 of the closest candidate. The candidate selection may also use other criteria instead of the closes distance, e.g., the largest distance. Note that the final number of candidates (or the codebook size) may not be same as the number of clusters. The AP may combine some clusters, e.g., the clusters with a small number of Vs (or H’s) into one cluster. Alternatively, the AP may divide one cluster, e.g., the cluster with large number of Vs (or H’s) into multiple clusters.

[0116] In one embodiment, the AP may indicate the clustering criteria, e.g., D_£ shown in FIG. 8, to the beamformee. In other words, the clustering criteria may need to be exchanged between the AP and non-AP STAs that perform clustering. The beamformee may, however, use a distance (e.g., clustering distance metric) or metric alternative to D_e, when it selects the CSI candidate vector to report. As an example, using the same D_e may result in high computational complexity for the beamformee to choose the closest candidate, which may result in additional delay of the CSI feedback. In such a condition, the beamformee may choose an alternative way to calculate the closest candidate which may not be as computationally expensive, and may or may not result in optimal candidate selection.

[0117] An example feedback procedure is shown in FIG. 9. Again, the distance metric D_e may not same as the what AP indicates. The metric used by the beamformee may be a variant version of the D_£ signaled by the AP. However, the non-AP STA may need to feedback the D_£ it used in the candidate selection phase. After the AP receives the index of the closest candidate to represent the current channel, the AP may select the corresponding candidate from the index. Depending on the representation of the candidates, the AP may then convert the candidate matrix to the corresponding steering matrix and start the transmission of data.

[0118] In one embodiment, after the AP determines the Al ML CSI related parameters, it may transmit an AIML CSI Parameter Element which includes AIML CSI related operation parameters.

[0119] Additional signaling may be used in certain embodiments. In one embodiment, the clustering criteria may be different from the criteria used in the candidate selection. In this case, the clustering criteria and/or the candidate selection criteria may need to be signaled between AP and non-AP STAs. For example, (1- General Cosine Similarity) may be used as the clustering criteria, which is used to determine the candidate beamforming matrices, i.e., Vs in the candidate set. After the candidate set is settled, when the beamformee feed backs the index of the current V matrix from the candidate set, the beamformee may use a different criterion, e.g., General Cosine Similarity (GCS) to determine which index of the candidate matrix should be fed back. If GCS is used in the candidate matrix selection in the beamformee side, the largest value of GCS between the beamforming matrix V and a candidate matrix V_candidate may imply this candidate matrix V_candidate is the best representation of the beamforming matrix V. The beamformee may feedback the corresponding index of the candidate matrix V_candidate.

[0120] Various embodiments relate to Procedures of AIML Enabled CSI Clustering. In one embodiment, before the AIML enabled CSI compression model is unified between AP and non-AP STAs, the AP may request the non-AP STA to indicate AIML model parameters used in its local CSI algorithm training, e.g., clustering inputs (features of clustering), clustering method, clustering criteria, the representation of the cluster centroid, the CSI candidate set form (or codebook form), the complexity of the clustering model, the testing accuracy (e.g., KPI value for the AIML CSI algorithms), etc. The AP may use the complexity number indicated by STAs to determine what CSI clustering method may be used for a uniform CSI compression model. Note that the AIML enabled CSI compression algorithm may not be limited to the clustering method described, as it may be applicable to any Al ML enabled CSI compression scheme or any CSI compression scheme. [0121] FIG. 10 shows an example method of AIML CSI compression model parameters exchange according to certain embodiments. In this example, the AP 1001 first transmits S1002 a request to a non-AP STA or multiple non-AP STAs 1003 to indicate AIML related parameters used in the non-AP STA CSI compression model training These parameters may include clustering method, clustering inputs (features of clustering), clustering criteria, representation of the cluster centroid, CSI candidate form (e.g., codebook form), complexity indicator and/or testing accuracy for indicated CSI compression algorithm(s). Upon reception of this request, the non-AP STA(s) may respond S1004 with the following parameters per request, e.g., clustering inputs (features of clustering), clustering method, clustering criteria, representation form of the cluster centroid, CSI candidate form (codebook form), complexity indicator of a given AIML enabled CSI compression algorithm(s) (e.g., clustering algorithm(s)), testing accuracy for the indicated algorithm(s) (e g., a KPI value of the corresponding algorithm). The non-AP STA may also indicate a preference of CSI compression algorithm it may want to use. After the AP collects the requested information from non-AP STAs, the AP may broadcast or unicast S1006 the unified AIML CSI model, e.g., clustering method, clustering inputs format, clustering criteria, representation form of the cluster centroid, CSI candidate form (codebook form) and/or any indicator/algorithm for AIML CSI compression the AP instructs non-AP STAs to use going forward.

[0122] In one embodiment, referring to FIG. 11 , the AP may use a trigger frame to request non-AP STA(s) to send back their AI L CSI compression algorithm related parameters. This trigger frame may be a new type of trigger frame or use any existing trigger frame type. For example, a new type of trigger frame may be referred to as an Al ML Parameters Report Poll (APRP) Trigger frame. An example of Trigger Dependent User Info subfield format an APRP Trigger frame is shown FIG. 11. In this example, two subfields are defined in the trigger Dependent User Info subfield of APRP Trigger frame: an AIML Control Information subfield 1101 and an AIML Parameters subfield 1103. In one example, the AIML Control information subfield contains N octets, for example, N =1.

[0123] An example of AIML Control Information subfield is shown in FIG. 12. In this example, a KPI Present subfield 1201 indicates if the AP indicates the KPI value in the AIML Parameters subfield or not, e.g. it is set =1 to indicate the KPI value is present; otherwise, it is not present. When KPI Present subfield is =1 , a KPI value will be indicated in the AIML Parameters field (FIG. 11) and it may be used to represent that the AP requires the non-AP STA to provide the testing accuracy based on the KPI value indicated in the AIML Parameters field. The Clustering Inputs Request subfield 1203 of the AIML Control Information field of FIG. 12, indicates if the AP requires the non-AP STA to indicate the cluster inputs or not, e.g. it is set =1 to indicate if clustering inputs are required to send back by the non-AP STA; otherwise it indicates that clustering inputs are not required to be fed back by the non-AP STA. AIML Control Information subfield of FIG. 12 may further include a Clustering Method Request subfield 1205 to indicate whether the AP requires the non-AP STA to indicate the cluster method or not, e.g., it is set =1 to indicate if the clustering method is required to be sent back by the non-AP STA; otherwise it indicates that the clustering method is not required to feedback. The AP may also specify the cluster method in the AIML Parameters to request the non-AP STA to test the specified method used in the AIML CSI compression algorithm. The AP may determine which clustering criteria the beamformees may use to feedback the index. The Clustering Criteria Request subfield 1207 may indicate if the Clustering Criteria field are present in the AIML Parameters field. In one embodiment, the AP may allow the beamformees to choose the clustering criteria used for the AIML index-based CSI feedback, and the Clustering Criteria Request subfield 1207 may indicate that the beamformee is required to signal explicitly the used clustering criteria. The Cluster Centroid Representation Request subfield 1209 indicates whether the AP requires the non-AP STA to send back the cluster centroid Representation format, e.g., it is set =1 to indicate the AP requests the non-AP to send back the cluster centroid representation format; otherwise it indicates that the AP may not request the non- AP STA to send back the cluster centroid representation format. The CSI Candidate Form Request subfield 1211 indicates whether the AP requests the non-AP STA to send back the CSI candidate form used in the CSI compression algorithm or not, e.g., it is set =1 to indicate the AP requests to the non-AP STA to send back the CSI candidate form used in the CSI compression algorithm, otherwise it does not request the non-AP STA to send back. A CSI Algorithm Complexity Request subfield 1213 may be included to indicate whether the non-AP STA responds with a complexity indicator or not, e.g., it is set =1 to indicate the AP requests the non-AP STA to send back the quantized complexity value for a given CSI

- ZU - compression algorithm; otherwise, it indicates that the AP may not request the non-AP STA to send back the CSI complexity value. The Preference of CSI Compression Algorithm Request subfield may request the non-AP STA to send back the preferred CSI compression algorithm or not. The AIML CSI compression algorithm may not be limited to clustering methods only and can be any type of AIML algorithm, e.g. supervised learning or non-supervised learning or reinforcement learning. For example, clustering inputs may represent the AIML CSI compression model inputs; the clustering method may represent the AIML method used for CSI compression; the clustering criteria may represent the optimization criteria (or objective/loss function) used in the AIML CSI compressions, etc. Although an AIML CSI compression scheme is used as an example in the disclosed embodiments, this trigger frame may be also used for the collection of AIML parameters used in any AIML enabled algorithms.

[0124] In one method, the AIML reports transmitted by the non-AP STAs may be included in the modified EHT Compressed Beamforming/CQI frame Action field as indicated in Table 1 below. The non-AP STAs may attach these AIML parameters along with EHT CSI reports, e.g., EHT compressed beamforming/CQI frame. In the example of Table 1, the AIML CSI Compression Parameters Report is included in the EHT Compressed Beamforming/CQI frame Action field, e.g. it follows the legacy Compressed CSI/CQI report and is transmitted within the same PPDU as the Compressed CQI/CQI report.

Order Meaning

7 AIML CSI Compression Parameters Report

TABLE 1 : Example of Modified EHT Compressed Beamforming/CQI frame Action field

[0125] Alternatively, AIML CSI Compression Parameters may be provided in response to an AP trigger frame by STAs by including a separate EHT Action field, e.g., AIML CSI Compression Model as indicated in Table 2 below.

Order Meaning

1 AIML CSI Compression Parameters

TABLE 2: Example of Modified EHT Action field

[0126] FIG. 13 is a message sequence diagram depicting an example method for AIML CSI Compression Parameters Reporting. In this example, an AP 1302 transmits the APRP Trigger frame 1301, or other similar functioning frame, which requests non-AP STAs 1304, 1306 to send back the AIML parameters indicated by the APRP Trigger frame. In this example STA1 and STA2 are shown but there is no limit on a number of participating STAs. Upon reception of the APRP Trigger frame, the non-AP STAs, e.g., non-AP STA1 1304 and non-AP STA2 1306, respond by including the requested AIML CSI compression related parameters 1303, 1305, for example, in the AIML CSI Compression Parameters Action field (e.g., as indicated in Table 2 above) or other similarly functioning messaging element, field or subfield. To this end, the terms “message,” “element,” “field” or “subfield” as used in this disclosure, are not intended as descriptors limiting any format or structure, and may be interchangeably used.

[0127] Various disclosed embodiments may relate to Methods to Enable AIML Based MU-MIMO CSI Feedback. In one embodiment, the beamformer (e.g., AP) may classify the beamformees (e.g., non-AP STAs) into multiple groups. The classification may depend on multiple criteria, e.g., CSI feedback from beamformees, channel quality indicator (CQI) report, signal interference to noise ratio (SI NR) report, modulation coding scheme (MCS) measurement, etc. When the beamformer transmits a NDP to multiple beamformees (e.g., using MU-MIMO transmission), it may indicate to the beamformees a group of STAs that each beamformee is paired with, e.g. by Group ID. For example, the Group ID can be indicated in the NDPA STA Infor field, or Trigger frame. When a beamformee sends back the CSI reports, it may include the group ID in report, e.g., in an EHT MIMO Control field. In the training phase where the candidate set is being generated, the beamformee may include the legacy CSI reports, which may include compressed beamforming, CQI report and MU SINR report, along with the Group ID in the CSI report. The beamformer may use the CSI reports and the Group ID information to obtain the corresponding candidate set for MU-MIMO or/and the received SINR information for each spatial stream given the paired STAs assigned in the MU-MIMO transmission. In other words, the candidate set may be changed with pairing STAs in the MU-MIMO transmission. Furthermore, the MCS assignment may change with pairing STAs in MU- MIMO transmission. It should be noted that the NDP MU-MIMO transmission may be precoded or non-precoded.

[0128] Referring to FIG. 14, in one example embodiment, two rounds of CSI reporting for MU-MIMO transmission may be utilized. The first round may be the legacy CSI reporting or AIML enabled CSI reporting, which includes SU-MIMO reports only. In the 2nd round, a precoded MU-MIMO NDP is transmitted by the AP. The AP may request the non-AP STA to indicate the differential CSI reports in this round. The differential CSI report may indicate the CSI difference between this round and the first round. In one example, the differential CSI report may include the beamforming matrix V difference, e.g., the difference between the reported V matrix in the first round and the 2nd round, the CQI (e.g., per-RU average SINR) difference between the first round and the 2nd round. FIG. 14 depicts an example of two-round enhanced EHT sounding operation for MU-MIMO. In this example, in the first round, the indexed based CSI reports are requested by the AP. Upon reception of such a request from the AP, non-AP STA1 and non-AP STA2 transmit back the index based CSI reports, for example 8x1 CSI report (one spatial stream). At a subsequent time, initiating the 2^nd round, the AP may use an enhanced NDPA frame to request the differential CSI reports from non-AP STA1 and non-AP STA2. In an example of this enhanced NDPA, a request for differential CSI reports may be included in the STA Info field of the enhanced NDPA frame. This request of differential CSI reports may, al ternati vely/additionally , be included in a separate trigger frame as shown in FIG. 14. After transmission of the enhanced NDPA frame, the AP transmit the NDP via precoded MU-MIMO transmission to non-AP STA1 and non-AP STA2. Upon reception of the enhanced NDPA, MU-MIMO transmitted NDP and the trigger frame, non-AP STAs measure the observed channel of the designated spatial stream. For example, 8x2 MU-MIMO is transmitted, i.e., two spatial streams are transmitted to STA1 and STA2. STA1 is designated to receive spatial stream #1 and STA2 is designated to receive spatial stream #2. Both STAs first measure and compute the beamforming matrix, the average SI NR for each spatial stream and/or CQI observed at this round. Then they further compute the difference between the beamforming matrix and/or the SI NR measured in this round and the 1st round and/or the difference between the beamforming CQI value measured in this round and the 1st round. The STAs include these difference in differential CSI reports, specifically the STA may indicate the additional interference present on this round. The interference may include inter-stream interference or interference from other DL transmissions. The STAs may also include the delta SI NR per subcarrier (which compares the SINR value on each subcarrier and the average SINR per spatial stream) in this round. It is noted that in the first round of CSI report, the report may be in any form, e.g., legacy CSI report or Al ML enabled CSI report. In the differential CSI report, the reporting may be an AIML enabled CSI differential report or a non-AIML enabled CSI differential report. Note that the 2^nd round DL transmission may be precoded or non-precoded MU-MIMO transmission. The 2^nd round sounding may appear more frequently than the first round.

[0129] Additionally, the two round sounding reports may be applied to SU-MIMO or CQI reports as well. In these case, the first round is to get the CSI and/or CQI reports for SU-MIMO transmission, which may use AIML or non-AIML enabled CSI and/or CQI reports. In the 2nd round, only the differential information is reported, e.g., the difference of the V matrices between the first round and the 2nd round, the difference of average SNR (or SINR) for each spatial stream between the first round and the 2nd round, the difference of CQI between the first round and the 2nd round, the additional inference present in the 2nd round, etc.

[0130] FIG. 15 depicts an example of multi-round EHT sounding operation for CQI report for SU-MIMO. In this example, the legacy CQI reports are solicited from STAs in the first round. In the 2nd round, the differential CQI reports are solicited by the AP. The solicited STAs transmit back the differential CQI report, which may include the differential SNR values represented by the difference of the average SNR per RU between the current round and the 1st round legacy CQI report, and/or the additional interference present in the current measurement. In the 3rd round the differential CQI reports are again solicited by the AP, which is some time after the 2nd round (differential CQI report). Per request from the AP, the STAs transmit back the differential CQI report, which may include the different SNR values represented by the difference of the average SNR per RU per spatial stream between the current round and the 1st round legacy CQI report, and/or the additional interference present in the current measurement compared with the 1st round. Alternatively, it may report the different SNR values between the average SNR per RU per spatial stream of the previous round (e.g., 2nd round) and the current round, and/or the additional interference present in the current measurement compared with the previous round (e.g., 2nd round). Please note the different SNR values presented in the Differential CQI report may be used in other forms.

[0131] FIG. 16 shows another example embodiment of a two-round enhanced EHT sounding procedure for MU-MIMO. In the second round CSI report, the AP or beamformer may transmit a NDP PPDU without precoding. The intended beamformees (e.g., Non-AP STA1 and Non-AP STA2 in the figure) may save the feedback information from the first round CSI report, and then calculate the difference between the feedback information and determine more accurate measurements. For example, for subcarrier group /, the beamformee (BFee) may get estimated channel H (here the subscript /' is omitted for simplicity). The BFee may perform singular value decomposition (SVD), so that H = USV^H. The BFee may use an index-based feedback for the 1st round, and fed back indexing to the AP may be used to recover an eigen vector matrix V, which may not be exactly the same as V, but close to it. For the 2nd round feedback, the BFee may perform channel estimation ft and SVD, and derive P. The BFee may calculate the difference D(V^H, 7) or £>( V). In one method, D(A, B) may be the correlation between the two matrices, i.e., D = D (A, B) = AB. The BFee may perform SVD and Givens decomposition on D matrix and feedback the resulting Givens angels to the beamformer (BFer). Note the 2nd round of enhanced NDPA transmission and NDP transmission may be omitted in some scenarios. In that case, V = V in above-mentioned example.

[0132] Additional embodiments for Al ML Training Algorithms are disclosed. In one embodiment, a Covariance Matrix Based Algorithm may be used. An AP may collect large amounts of channel state information from a database to train Nc centroids. While an AP is used to illustrate the algorithm of the embodiments, the AP may be replaced by controller, or other type of device which may be capable to perform AIML training.

[0133] APs and STAs which support covariance matrix based algorithm may indicate the same in a capability element/field in a management/control frame. In the training phase, the AP may follow below procedures to train the centroids:

[0134] For each MIMO channel state information H observed in frequency domain for one or more subcarriers in the database (with size N_r x N_t, where Nr is the number of receive antennas and Nt is the number of transmit antennas), the AP may construct a covariance matrix K in one of the manners.

[0135] (i) The AP may perform singular value decomposition (SVD) on H, so that H = USV^H, where U (with size

N_r x W_r) and V (with size N_t x N_t) are Unitary matrices and S is a N_r x N_t rectangular diagonal matrix. The covariance matrix is defined as: K = VV^H or

[0136] (ii) The covariance matrix is defined as K = H x H^H /C, where C is a normalized factor. For example, C = \\H \\p could be a p-norm, or other kind of norm.

[0137] The AP may perform k-means clustering on all of the covariance matrices in the database. Euclidean distance or squared Euclidean distance may be used for k-means clustering. After the k-means algorithm, Nm centroids may be selected and the AP may send the Nm centroids to its STAs so that both beamformer and beamformee know the centroids.

[0138] In the beamforming sounding phase, A beamformer may indicate in a NDP Announcement frame the algorithm used to train the centroids. For example, an AIML Training Type field may be carried in the NDPA frame, or other frame, transmitted by the beamformer to announce the beamforming training/sounding requirements. One value for the AIML Training Type field may be used to indicate Covariance matrix based k-means algorithm with Euclidian distance or square Euclidian distance may be used for database training. Note this information is used for the beamformee to select a centroid index properly to feedback to the beamformer if more than one AIML training algorithms may be allowed. Alternatively, the configuration may select and support only one AIML index based training algorithm. In this case, signaling for the AIML Training Type may be omitted.

[0139] The beamformer may transmit one or more sounding frames, such as NDP PPDUs. On reception the NDPA and NDP frames, a intended beamformee may prepare the beamforming feedback index selection. Based on the sounding frame, the beamformee may estimate one or more MIMO/SISO channel matrices where / may be the subcarrier group index. Note, in NDPA frame, the beamformer may request one CSI index feedback for a subcarrier group. The beamformee may construct the covariance matrix K_L.

[0140] In one method, the beamformee may perform SVD on 'H_l, 'H_l = The beamformee may obtain the covariance matrix K_t = VjV^ . In another method, the beamformee may determine the covariance matrix using the following equation:

[0141] .where Cj is a normalized factor. For example, C_t — \\Hi ||^ could be a p-norm, or other kind of norm.

[0142] The beamformee compares K_t with all the centroids [#i, , K_Nm], and selects a centroid index n, where centroid n, is the closest node to K_n under certain criteria. The criteria may be minimizing the Euclidean distance or squared Euclidean distance, for example, using the following equation:

[0143] The beamformee may feedback index n, for subcarrier group /.

[0144] When the training dataset is composed by the covariance matrices, Euclidean distance or squared Euclidean distance may be used to maximize the beamforming gain or capacity. Using Euclidean distance or squared Euclidean distance as criteria or intermediate KPI is equivalent to Generalized cosine similarity (GCS) or squared GCS (SGCS) when a MISO channel is considered or single data stream transmission is considered. Note, GCS/SGCS is defined for vectors and it requires extra work to extend it to matrix case which may serve MIMO channel or multiple data stream transmissions. Using Euclidean distance or squared Euclidean distance on covariance matrices can be easily extended to MIMO case or multiple data stream transmission case. [0145] The following is a proof of using Euclidean distance or squared Euclidean distance as criteria or intermediate KPI is equivalent to Generalized cosine similarity (GCS) or squared GCS (SGCS) when a MISO channel is considered or single data stream transmission is considered. The covariance matrix of the channel matrix H is calculated according to K = H^HH, where the superscript denotes the Hermitian of the matrix. The channel matrix H can be decomposed using singular value decomposition (SVD) as H — USV^H , where U and V are unitary matrices and S is a diagonal matrix with the corresponding singular values. Substituting this back into the covariance matrix gives K = ( JSV^H)^H (USV^H) = VS^H U^H USV^H = VS²V^H, where U^HU = I is a unitary matrix, and the matrix S² contains the square of the singular values at the diagonal. If the size of H is N x M, the size of V is M x N (only N columns of V picked corresponding to the highest singular values), and the size of K will be M x M. The covariance matrix K is essentially the outer product of V and V^H scaled by the square of the singular values. For example, consider a V of size M x 1 (N = 1), given by V = v₂, ... , v_M]^r. The covariance matrix in this case is given by:

[0146] .where s is the singular value. For K-means clustering using this method, the squared Euclidean distances are computed between the covariance matrix of a channel H in the dataset and the cluster centroid (initialized by picking covariance matrices randomly from the dataset).

[0147] Let, K_a be a covariance matrix picked from our dataset, and K_b be the covariance matrix representing a cluster centroid. Then the squared Euclidean distance between them (d_cov) after each covariance matrix is normalized by its two-norm can be represented by equation:

= 2 - 2\V_a ^HV_b \² Eq. 14 d_cov = 2(1 - \V_a ^HV_b \²) Eq. 15

[0148] In the above equations, x* represents the conjugate of the complex number x.

[0149] Embodiments for Separate Clustering of p and ip Angles are disclosed in which, in one embodiment, separate candidate vectors may be obtained to represent the feedback angle indices for the ( and ip angles. In such a method, there may be two different datasets, one with the <p angle index data and another with the ip angle index data. These datasets may then be input into a clustering algorithm (e.g. K-means clustering as previously discussed) separately to obtain the candidate cluster sets for <p and ip angle index data. An example of this separate clustering is shown in FIG. 16. In this example, vectors that contain p angles are input of one Al ML clustering algorithm, i.e., AIML Clustering Algorithm- 1, and vectors that contain ip angles are input of another AIML clustering algorithm, i.e., AIML Clustering Algorithm-2. These two algorithm may be the same or different. The output of the clustering algorithms of FIG. 16 are the centroids of a given number of clusters. The centroids of the clusters are the candidates for use. It is noted that, while this is an example using AIML clustering algorithm(s). The algorithm may be any type of AIML algorithms, e.g., unsupervised learning, supervise learning or reinforcement learning or the like.

[0150] Performing separate clustering on the two different types of angles, as shown in FIG. 17, may allow flexibility to assign different number of feedback bits (or candidate vectors) for the ( and p angles. As an example, consider that the number of clusters for ( vectors is and the number of clusters for ip vectors is C₂. In such a case, the total number of bits required for each subcarrier group in this example is log₂ + log₂ C₂.

[0151] Embodiments for Density Enhanced K-Mean determinations are also disclosed. Referring to FIGs. 18-19, the frequency of use of each candidate in a candidate set may be analyzed to further improve the quality of the candidate set. As an example, during beamforming report generation, in a practical scenario, there may be instances where certain candidates are used frequently, whereas some other candidates are used rarely, based on the channel conditions in which the candidate set is being used. After such data is collected, the candidates in a given set may be arranged in the order of the frequency of their use. A certain number of low frequency candidates may be dropped altogether in such a case, resulting in even lower number of bits required for the beamforming feedback in the beamforming report. The number of low frequency candidates to be dropped in this example may be implementation dependent. [0152] In one embodiment, as shown in FIG. 18, multiple candidates may be obtained from the subsets of the dataset that correspond to the high frequency candidates (or centroids). As an example, let N_r be the number of high frequency candidates being picked (while the others are dropped from the candidate set). Within the subset of the data corresponding to each of these N_r candidates, N_s new candidates may be found. The original candidate set may be replaced by N_r x N_s new candidates In this specific example, N_s is chosen such that N_r x N_s = N_m, where N_m is the original size of the candidate set.

[0153] In one embodiment, referring to FIG. 19, /V_0.A, number of low frequency candidates may be chosen. Further, the subsets of the dataset represented by these candidates may be combined. On this combined dataset, clustering may be performed so that N_e number of new candidates are generated (N_e < N_low). Then N_high number of high frequency candidate vectors may be chosen. From each of the subsets of the dataset representing these candidate vectors, N_s new candidates may be obtained. All the variables mentioned in this method may be implementation dependent. In this example, N_u represents the number of candidates that were unchanged and N_s = 2.

[0154] In another embodiment, for each of the N_low number of low frequency clusters being considered, the nearest cluster (that doesn't belong in the f|_0W clusters) may be found, and the two subsets may be combined. From this combined subset of data, one or multiple candidates may be obtained. Correspondingly, from the subsets of the dataset representing the W_high number of high frequency candidates, one or multiple candidates may be obtained.

[0155] Embodiments for Clustering Considering Effective Distance Between Two vectors are further disclosed. In one embodiment, the distance (or effective difference) calculated between the angles may be considered while clustering. As an example, consider any <p angle index in a feedback vector. While clustering, the algorithm may calculate the squared Euclidean distance between a vector in consideration and all the cluster centroids.

[0156] The squared Euclidean distance is the sum of squared difference between all the angle indexes in the feedback vectors. Let the ith <f> angle index in the feedback vector a be cp_{i a} and the ith <f> angle index in feedback vector b be (f>_{i b}. the difference between these two angle indexes is given by the equation: <Pi = (<Pi,a ~ <Pi,b) Eq. 16

[0157] Mathematically, it may be considered that the effective distance (D_eff) between these two angles is given by: e_ff = A< >; = 2n — A< >; Eq. 17

[0158] FIG. 20 illustrates using a phasor representation how A< >; and 2rr - A< >, are a same distance away from 0 making their effective distance to be the same. This effective distance between all <f> angles may be taken into account while clustering to generate the candidate vectors. [0159] Embodiments for Candidate Selection Using K Nearest Neighbors Algorithm (k-NN) are further disclosed where, in one embodiment, the supervised learning algorithm k-NN may be used for candidate selection while generating the beamforming report. In a generated candidate set, each candidate may represent a subset of the dataset over which clustering is performed. This representation may be used as the labeling required for the k-NN algorithm. After generating the feedback report, the non-AP STA may compute k nearest neighbors in the dataset. The non-AP STA may then choose the candidate representing the majority of the k nearest neighbors. The number k in this instance may be implementation dependent.

[0160] Embodiments for Clustering with Lower Quantization Order for ip are further considered. In one embodiment, using a high quantization order for ip angles may be considered to be detrimental to the accuracy while clustering. In such an instance, the quantization order used to quantized p angles may be reduced.

[0161] According to additional embodiments, a serialized V method may be used where one or more data stream sounding training may be utilized. In an example training phase, the AP (or other device/entity) may use the following methods to train the centroids.

[0162] For each MIMO channel state information H observed in frequency domain for one or more subcarriers in the database (with size N_r x N_t, where Nr is the number of receive antennas and Nt is the number of transmit antennas).

[0163] The AP may perform singular value decomposition (SVD) on H, so that H — USV^H , where U (with size N_r x N_r) and V (with size N_t x N_t) are Unitary matrices and S is a N_r x N_t rectangular diagonal matrix. Denote V = [v_1; and each v_k (k=1.„„Nt) is a N_r x 1 column vector. Note though there are Nt columns for V matrix, only the first Ns columns are meaningful. Ns could be the rank of the H matrix or the number of data streams the AP may want to transmit. In this way, the V matrix may be shortened to size N_t x N_s.

[0164] The AP may serialize the V matrix by: the size of SV is N_r N_s x 1 Eq. 18

[0165] The AP may perform k-means clustering on all of the SV vectors in the database. In one method, an orthonormalization process, generalized cosine similarity (GCS) may be used as criteria for k-means clustering. In one method, Euclidean distance or squared Euclidean distance may be used for k-means clustering.

[0166] After the k-means algorithm, Nm centroids may be selected. In one method, each centroid serialized V vector, denoted CSVk may be determined as:

[0167] and size is N_rN_s x 1 may be the mean vector of the SV vectors (i ,e. , each element in the CSVk is the mean of the elements of vectors in the cluster) in the cluster. A centroid V matrix may be obtained by CV_k = [v_kl, , v_kN ] with size N_t x N_s. Matrix CV may not be a unitary matrix anymore due to the average operation. An orthonormalization process, e.g., Gram-Schmidt process, may be used to make matrix CV unitary again. The AP may send the Nm centroids to its STAs so that both beamformer and beamformee know the centroids.

[0168] A sounding procedure for these embodiments may be similar as described before. In one example embodiment, the beamformee may estimate one or more Ml MO/SISO channel matrices Hl, where i may be the subcarrier group index.

[0169] The beamformee may perform SVD on T-T_lt 'Hl = U_lS_lV^H _i. The beamformee may compare V_t with the centroid V matrices [CV ... , CV_Wm], and select a centroid index ni where centroid ni is the closest node to V_t under certain criteria. In one method, GCS may be used as the criteria. In one method, Euclidean distance or squired Euclidean distance may be used as the criteria. The beamformee may then feedback index ni for subcarrier group /.

[0170] Although the features and elements are described in the preferred embodiments in particular combinations, each feature or element can be used alone without the other features and elements of the preferred embodiments or in various combinations with or without other features and elements of the preferred embodiments.

[0171] Although the solutions described herein are discussed in reference to IEEE 802.11 specific protocols, it is understood that the solutions described herein are not restricted to this scenario and are applicable to other wireless systems as well. Although SIFS is used to indicate various inter frame spacing in the examples of the designs and procedures, all other inter frame spacing such as RIFS, AIFS, DIFS or other agreed time interval could be applied in the same solutions. Although specific examples are described or referenced in the drawing, the numbers and frequencies of RBs/channels/bandwidth utilized may vary. Further, while specific bits are used to signal in-BSS/OBSS as example, other bit(s) may be used to signal this information.

[0172] Although features and elements are described above in particular combinations, one of ordinary skill in the art will appreciate that each feature or element can be used alone or in any combination with the other features and elements. In addition, the methods described herein may be implemented in a computer program, software, or firmware incorporated in a computer-readable medium for execution by a computer or processor. Examples of computer-readable media include electronic signals (transmitted over wired or wireless connections) and computer-readable storage media. Examples of computer-readable storage media include, but are not limited to, a read only memory (ROM), a random access memory (RAM), a register, cache memory, semiconductor memory devices, magnetic media such as internal hard disks and removable disks, magneto-optical media, and optical media such as CD-ROM disks, and digital versatile disks (DVDs). A processor in association with software may be used to implement a radio frequency transceiver for use in a WTRU, UE, terminal, base station, RNC, or any host computer.

Claims

CLAIMS What is Claimed:

1. A method for an access point, AP, the method comprising: sending, to one or more stations, information indicative of a request for channel state information, CSI; receiving, from at least one station, information indicative of CSI; clustering received CSI to obtain indexed information for a set of clusters; and sending information indicative of the indexed information for the set of clusters to at least one station to be used for CSI feedback procedure.

2 The method of claim 1 , wherein the information indicative of CSI comprises artificial intelligence machine learning, AIML, parameters respectively used by the at least one station in the set for local CSI AIML training.

3 The method of claim 2, wherein the AIML parameters include at least one of clustering method, clustering input, clustering criteria, representation of a cluster centroid, CSI candidate set form, complexity number of a used clustering model, and testing accuracy.

4 The method of claim 3, further comprising determining a clustering method based on the complexity number.

5 The method of claim 1 , wherein the receiving comprises receiving an indication of a preference of CSI compression algorithm.

6 The method of claim 1 , wherein the indexed information for the set of clusters includes at least one of clustering method, clustering input, clustering criteria, representation of a cluster centroid, CSI candidate set form, complexity number of a used clustering model, and testing accuracy.

7 The method of claim 1 , wherein the request is sent as a trigger frame.

8 The method of claim 2, wherein: clustering a plurality of the AIML parameters comprises identifying a plurality of centroids; and wherein sending information indicative of the indexed information for the set of clusters comprises sending the identified plurality of centroids for use in AIML CSI feedback.

9 The method of claim 8, wherein identifying each the plurality of centroids includes performing singular value decomposition (SVD) on observed multiple input multiple output (MIMO) CSI (H) to derive a CSI matrix (V), serializing V to a column vector (SV) based on a number of transmit antennas and a rank of H or number of transmit data streams, and clustering centroids based on SV vectors.

10. The method of claim 9, further comprising: sounding one or more multiple input multiple output (Ml MO) channels to determine CSI for a number of receive antennas and the number of transmit antennas; and performing the SVD on the determined CSI.

11. The method of claim 9, wherein clustering centroids comprises applying a K-means clustering algorithm on centroid serialized V vectors.

12. The method of claim 1 , further comprising: sending, to at least one station, information indicative of a request for information regarding a subset of the set of clusters; and receiving, from the at least one station, cluster indices corresponding to the subset of the set of clusters.

13. An access point, AP, comprising at least one processor configured to: send, to one or more stations, information indicative of a request for channel state information, CSI; receive, from at least one station, information indicative of CSI; cluster received CSI to obtain indexed information for a set of clusters; and send information indicative of the indexed information for the set of clusters to at least one station to be used for CSI feedback procedure.

14. The access point of claim 13, wherein the information indicative of CSI comprises artificial intelligence machine learning, Al ML, parameters respectively used by the at least one station in the set for local CSI AIML training.

15. The access point of claim 14, wherein the AIML parameters include at least one of clustering method, clustering input, clustering criteria, representation of a cluster centroid, CSI candidate set form, complexity number of a used clustering model, and testing accuracy.

16. The access point of claim 15, wherein the processor is further configured to determine a clustering method based on the complexity number.

17. The access point of claim 13, receive, from at least one station, information indicative of CSI comprises receiving an indication of a preference of CSI compression algorithm.

18. The access point of claim 13, wherein the indexed information for the set of clusters includes at least one of clustering method, clustering input, clustering criteria, representation of a cluster centroid, CSI candidate set form, complexity number of a used clustering model, and testing accuracy.

19. The access point of claim 13, wherein the request is sent as a trigger frame.

20. The access point of claim 14, wherein the at least one processor is further configured to: cluster a plurality of the Al ML parameters comprises identifying a plurality of centroids; and wherein send information indicative of the indexed information for the set of clusters comprises sending the identified plurality of centroids for use in AIML CSI feedback.

21 . The access point of claim 20, wherein identifying each the plurality of centroids includes performing singular value decomposition (SVD) on observed multiple input multiple output (MIMO) CSI (H) to derive a CSI matrix (V), serializing V to a column vector (SV) based on a number of transmit antennas and a rank of H or number of transmit data streams, and clustering centroids based on SV vectors.

22. The access point of claim 21 , wherein the at least one processor is further configured to: sound one or more multiple input multiple output (MIMO) channels to determine CSI for a number of receive antennas and the number of transmit antennas; and perform the SVD on the determined CSI.

23. The access point of claim 21 , wherein clustering centroids comprises applying a K-means clustering algorithm on centroid serialized V vectors.

24. The access point of claim 13, wherein the at least one processor is further configured to: send, to at least one station, information indicative of a request for information regarding a subset of the set of clusters; and receive, from the at least one station, cluster indices corresponding to the subset of the set of clusters.

25. A method for a wireless station associated with an access point, the method comprising: receiving, from the access point, information indicative of a request for information about a set of channel state information, CSI, clusters; and sending, to the access point, requested cluster indices.

26. The method of claim 25, further comprising, prior to the receiving: receiving, from the access point, information indicative of a request for CSI; sending, to the access point, information indicative of CSI; and receiving, from the access point, information indicative of indexed information for the set of clusters to at least one station to be used for CSI feedback procedure.

27. The method of claim 26, wherein the information indicative of CSI comprises artificial intelligence machine learning, AIML, parameters respectively used by the at least one station in the set for local CSI Al ML training.

28. A wireless station configured for association with an access point, the wireless station comprising at least one processor configured to: receive, from the access point, information indicative of a request for information about a set of channel state information, CSI, clusters; and send, to the access point, requested cluster indices.

29. The wireless station of claim 28, wherein the at least one processor is further configured to, prior to the receive: receive, from the access point, information indicative of a request for CSI; send, to the access point, information indicative of CSI; and receive, from the access point, information indicative of indexed information for the set of clusters to at least one station to be used for CSI feedback procedure.

30. The wireless station of claim 29, wherein the information indicative of CSI comprises artificial intelligence machine learning, AIML, parameters respectively used by the at least one station in the set for local CSI AIML training.