Beamforming Methods And Methods For Using Beams Patent Application (2024)

U.S. patent application number 17/473506 was filed with the patent office on 2021-12-30 for beamforming methods and methods for using beams.This patent application is currently assigned to INTERDIGITAL PATENT HOLDINGS, INC.. The applicant listed for this patent is INTERDIGITAL PATENT HOLDINGS, INC.. Invention is credited to Monisha Ghosh, Hanqing Lou, Robert L. Olesen, Oghenekome Oteri, Pengfei Xia.

BEAMFORMING METHODS AND METHODS FOR USING BEAMS

Abstract

A method, apparatus, system, and computer readable medium may beused to perform beamforming. The method may include a firstcommunication device sending a first plurality of beamformingtraining frames to a second communication device using a firstbeamforming weight vector; the first communication device receivingfrom the second communication device a second beamforming weightvector; and the first communication device sending a secondplurality of beamforming training frames to the secondcommunication device using the second beamforming weight vector.The apparatus, method, system, and computer readable media may usespatial diversity with beam switching, spatial diversity with asingle beam, weighted multipath beamforming training, single userspatial multiplexing, and beamforming training for beam divisionmultiple access (BDMA).

Related U.S. Patent Documents

Claims

1. A method performed by a first station (STA) comprising aplurality of antennas, the method comprising: partitioning theplurality of antennas into at least a first group of antennas and asecond group of antennas, wherein the first group of antennas isassociated with a first sector to a second STA, and the secondgroup of antennas is associated with a second sector to the secondSTA; transmitting, to the second STA, a plurality of beamformingtraining frames using the first group of antennas and the secondgroup of antennas; receiving, from the second STA, a firstbeamforming weight vector for sending signals on the first group ofantennas; and receiving, from the second STA, a second beamformingweight vector for sending signals on the second group of antennas.

Description

CROSS REFERENCE TO RELATED APPLICATIONS

[0001] This application is a continuation of U.S. patentapplication Ser. No. 14/441,237, filed May 7, 2015, which is anational stage entry of PCT/US2013/69265, filed Nov. 8, 2013 whichclaims the benefit of U.S. Provisional Application No. 61/724,679filed on Nov. 9, 2012, which are incorporated by reference as iffully set forth herein.

BACKGROUND

[0002] Some wireless communication networks support operation atvery high and even extremely high carrier frequencies such as 60GHz and millimeter wave (mmW) frequency bands. These extremely highcarrier frequencies may support very high throughput such as up to6 gigabits per second (Gbps). One of the challenges for wirelesscommunication at very high or extremely high carrier frequencies isthat a significant propagation loss may occur due to the highcarrier frequency. As the carrier frequency increases, the carrierwavelength may decrease, and the propagation loss may increase aswell.

[0003] At mmW frequency bands, the propagation loss may be severe.For example, the propagation loss may be on the order of 22 to 27dB, relative to that observed in either the 2.4 GHz, or 5 GHzbands. Since the available spectrum is limited, however, and sinceusers continue to demand more bandwidth, there is a need foreffectively using very high and extremely high carrier frequenciesfor communication networks.

BRIEF DESCRIPTION OF THE DRAWINGS

[0004] FIG. 1A is a system diagram of an example communicationssystem in which one or more disclosed embodiments may beimplemented;

[0005] FIG. 1B is a system diagram of an example wirelesstransmit/receive unit (WTRU) that may be used within thecommunications system illustrated in FIG. 1A;

[0006] FIG. 1C is a system diagram of an example radio accessnetwork and an example core network that may be used within thecommunications system illustrated in FIG. 1A;

[0007] FIG. 1D is a diagram of an example WLAN with an AP and STAforming a BSS, and beamforming with multipath channels;

[0008] FIG. 1E is a diagram of another example WLAN with an AP andSTA forming a BSS, and beamforming with multiple channels;

[0009] FIG. 2 is a diagram of an example of a method using two STAsto perform a multi-path beamforming method;

[0010] FIG. 3 is a diagram of an example of one iteration of themulti-path beamforming method using BRP transactions;

[0011] FIG. 4 is a diagram of an example frame format of a BRPpacket;

[0012] FIG. 5 is a diagram of an example format of a BRP frameAction field;

[0013] FIG. 6 is a diagram of an example modified channelmeasurement feedback element;

[0014] FIG. 7 is a diagram of an example AP configured to perform atransmission using full size beamforming with STBC;

[0015] FIG. 8 is a diagram of an example AP configured to perform atransmission using partial size beamforming with STBC;

[0016] FIG. 9 is a diagram of an example transceiverarchitecture;

[0017] FIG. 10 is a diagram of another example transceiverarchitecture;

[0018] FIG. 11 is a diagram of an example beam division multipleaccess (BDMA) architecture;

[0019] FIG. 12 is a diagram of an example beamforming trainingmethod for BDMA;

[0020] FIG. 13 is a diagram of an example modified BRP procedure toimplement a multi-stage iterative beamforming training method forBDMA;

[0021] FIG. 14 is a diagram of an example PHY layer frameformat;

[0022] FIG. 15 is a diagram of an example beamforming trainingprocedure using Eigen-beamforming based spatial multiplexing wherethe communication devices may be configured to calibrate multipletransmit RF chains;

[0023] FIG. 16 is a diagram of an example beamforming trainingmethod for Type I devices and for Type II devices withoutcalibration;

[0024] FIG. 17 is a diagram of an example beamforming trainingmethod for beam sweep based spatial multiplexing for Type I deviceswith calibration between two TX chains;

[0025] FIG. 18 is a diagram of an example beamforming trainingmethod for beam sweep based spatial multiplexing Type II devicesand Type I devices without calibration;

[0026] FIG. 19 is a diagram of an example modified FBCK-TYPEsubfield;

[0027] FIG. 20A is a diagram of an example PHY layer frameformat;

[0028] FIG. 20B is a diagram of another example PHY layer frameformat;

[0029] FIG. 20C is a diagram of another example PHY layer frameformat;

[0030] FIG. 21 is a diagram of an example modified SSW trainingframes and sequence;

[0031] FIG. 22 is a diagram of an example SSWA frame format;

[0032] FIG. 23 is a diagram of an example early termination of theSLS training procedure; and

[0033] FIG. 24 is a diagram of an example multi-beam multi-DMGantenna SLS feedback method.

SUMMARY

[0034] A first communication device for beamforming may include aplurality of antennas and a processor. The processor may beconfigured to partition the antenna into at least a first group ofantennas and a second group of antennas. The processor may befurther configured to send a plurality of beamforming trainingframes to a second communication device using the first group ofantennas and the second group of antennas. The processor and/or areceiver may be configured to receive, from the secondcommunication device, a first beamforming weight vector for sendingsignals on the first group of antennas and to receive a secondbeamforming weight vector for sending signals on the second groupof antennas.

[0035] A method of beamforming training for beam division multipleaccess (BDMA) may include an AP transmitting Nt sequences modulatedusing Nt beamforming vectors. A first station may use a firstprevious beamforming vector to receive the Nt sequences anddetermine a first transmit beamforming weight from the AP to thefirst station based on the first previous beamforming vector andthe received Nt sequences. The first station may send thedetermined first transmit beamforming weight to the AP. A secondstation may use a second previous beamforming vector to receive theNt sequences and determine a second transmit beamforming weightfrom the AP to the first station based on the second previousbeamforming vector and the received Nt sequences. The secondstation may send the determined second transmit beamforming weightto the AP, and the AP may transmit one or more sequences modulatedbased on the first transmit beamforming weight and the secondtransmit beamforming weight.

[0036] A method and apparatus may be used for spatial diversitywith beam switching, spatial diversity with a single beam, weightedmultipath beamforming training, single user spatial multiplexing,and for reduced beamforming training overhead.

DETAILED DESCRIPTION

[0037] FIG. 1A is a diagram of an example communications system 100in which one or more disclosed embodiments may be implemented. Thecommunications system 100 may be a multiple access system thatprovides content, such as voice, data, video, messaging, broadcast,etc., to multiple wireless users. The communications system 100 mayenable multiple wireless users to access such content through thesharing of system resources, including wireless bandwidth. Forexample, the communications systems 100 may employ one or morechannel access methods, such as code division multiple access(CDMA), time division multiple access (TDMA), frequency divisionmultiple access (FDMA), orthogonal FDMA (OFDMA), single-carrierFDMA (SC-FDMA), and the like.

[0038] As shown in FIG. 1A, the communications system 100 mayinclude wireless transmit/receive units (WTRUs) 102a, 102b, 102c,102d, a radio access network (RAN) 104, a core network 106, apublic switched telephone network (PSTN) 108, the Internet 110, andother networks 112, though it will be appreciated that thedisclosed embodiments contemplate any number of WTRUs, basestations, networks, and/or network elements. Each of the WTRUs102a, 102b, 102c, 102d may be any type of device configured tooperate and/or communicate in a wireless environment. By way ofexample, the WTRUs 102a, 102b, 102c, 102d may be configured totransmit and/or receive wireless signals and may include userequipment (UE), a mobile station, a fixed or mobile subscriberunit, a pager, a cellular telephone, a personal digital assistant(PDA), a smartphone, a laptop, a netbook, a personal computer, awireless sensor, consumer electronics, and the like.

[0039] The communications systems 100 may also include a basestation 114a and a base station 114b. Each of the base stations114a, 114b may be any type of device configured to wirelesslyinterface with at least one of the WTRUs 102a, 102b, 102c, 102d tofacilitate access to one or more communication networks, such asthe core network 106, the Internet 110, and/or the networks 112. Byway of example, the base stations 114a, 114b may be a basetransceiver station (BTS), a Node-B, an eNode B, a Home Node B, aHome eNode B, a site controller, an access point (AP), a wirelessrouter, and the like. While the base stations 114a, 114b are eachdepicted as a single element, it will be appreciated that the basestations 114a, 114b may include any number of interconnected basestations and/or network elements.

[0040] The base station 114a may be part of the RAN 104, which mayalso include other base stations and/or network elements (notshown), such as a base station controller (BSC), a radio networkcontroller (RNC), relay nodes, etc. The base station 114a and/orthe base station 114b may be configured to transmit and/or receivewireless signals within a particular geographic region, which maybe referred to as a cell (not shown). The cell may further bedivided into cell sectors. For example, the cell associated withthe base station 114a may be divided into three sectors. Thus, inone embodiment, the base station 114a may include threetransceivers, i.e., one for each sector of the cell. In anotherembodiment, the base station 114a may employ multiple-inputmultiple output (MIMO) technology and, therefore, may utilizemultiple transceivers for each sector of the cell.

[0041] The base stations 114a, 114b may communicate with one ormore of the WTRUs 102a, 102b, 102c, 102d over an air interface 116,which may be any suitable wireless communication link (e.g., radiofrequency (RF), microwave, infrared (IR), ultraviolet (UV), visiblelight, etc.). The air interface 116 may be established using anysuitable radio access technology (RAT).

[0042] More specifically, as noted above, the communications system100 may be a multiple access system and may employ one or morechannel access schemes, such as CDMA, TDMA, FDMA, OFDMA, SC-FDMA,and the like. For example, the base station 114a in the RAN 104 andthe WTRUs 102a, 102b, 102c may implement a radio technology such asUniversal Mobile Telecommunications System (UMTS) Terrestrial RadioAccess (UTRA), which may establish the air interface 116 usingwideband CDMA (WCDMA). WCDMA may include communication protocolssuch as High-Speed Packet Access (HSPA) and/or Evolved HSPA(HSPA+). HSPA may include High-Speed Downlink Packet Access (HSDPA)and/or High-Speed Uplink Packet Access (HSUPA).

[0043] In another embodiment, the base station 114a and the WTRUs102a, 102b, 102c may implement a radio technology such as EvolvedUMTS Terrestrial Radio Access (E-UTRA), which may establish the airinterface 116 using Long Term Evolution (LTE) and/or LTE-Advanced(LTE-A).

[0044] In other embodiments, the base station 114a and the WTRUs102a, 102b, 102c may implement radio technologies such as IEEE802.16 (i.e., Worldwide Interoperability for Microwave Access(WiMAX)), CDMA2000, CDMA2000 1.times., CDMA2000 EV-DO, InterimStandard 2000 (IS-2000), Interim Standard 95 (IS-95), InterimStandard 856 (IS-856), Global System for Mobile communications(GSM), Enhanced Data rates for GSM Evolution (EDGE), GSM EDGE(GERAN), and the like.

[0045] The base station 114b in FIG. 1A may be a wireless router,Home Node B, Home eNode B, or access point, for example, and mayutilize any suitable RAT for facilitating wireless connectivity ina localized area, such as a place of business, a home, a vehicle, acampus, and the like. In one embodiment, the base station 114b andthe WTRUs 102c, 102d may implement a radio technology such as IEEE802.11 to establish a wireless local area network (WLAN). Inanother embodiment, the base station 114b and the WTRUs 102c, 102dmay implement a radio technology such as IEEE 802.15 to establish awireless personal area network (WPAN). In yet another embodiment,the base station 114b and the WTRUs 102c, 102d may utilize acellular-based RAT (e.g., WCDMA, CDMA2000, GSM, LTE, LTE-A, etc.)to establish a picocell or femtocell. As shown in FIG. 1A, the basestation 114b may have a direct connection to the Internet 110.Thus, the base station 114b may not be required to access theInternet 110 via the core network 106.

[0046] The RAN 104 may be in communication with the core network106, 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. Forexample, the core network 106 may provide call control, billingservices, mobile location-based services, pre-paid calling,Internet connectivity, video distribution, etc., and/or performhigh-level security functions, such as user authentication.Although not shown in FIG. 1A, it will be appreciated that the RAN104 and/or the core network 106 may be in direct or indirectcommunication with other RANs that employ the same RAT as the RAN104 or a different RAT. For example, in addition to being connectedto the RAN 104, which may be utilizing an E-UTRA radio technology,the core network 106 may also be in communication with another RAN(not shown) employing a GSM radio technology.

[0047] The core network 106 may also serve as a gateway for theWTRUs 102a, 102b, 102c, 102d to access the PSTN 108, the Internet110, and/or other networks 112. The PSTN 108 may includecircuit-switched telephone networks that provide plain oldtelephone service (POTS). The Internet 110 may include a globalsystem of interconnected computer networks and devices that usecommon communication protocols, such as the transmission controlprotocol (TCP), user datagram protocol (UDP) and the internetprotocol (IP) in the TCP/IP internet protocol suite. The networks112 may include wired or wireless communications networks ownedand/or operated by other service providers. For example, thenetworks 112 may include another core network connected to one ormore RANs, which may employ the same RAT as the RAN 104 or adifferent RAT.

[0048] Some or all of the WTRUs 102a, 102b, 102c, 102d in thecommunications system 100 may include multi-mode capabilities,i.e., the WTRUs 102a, 102b, 102c, 102d may include multipletransceivers for communicating with different wireless networksover different wireless links. For example, the WTRU 102c shown inFIG. 1A may be configured to communicate with the base station114a, which may employ a cellular-based radio technology, and withthe base station 114b, which may employ an IEEE 802 radiotechnology.

[0049] FIG. 1B is a system diagram of an example WTRU 102. As shownin FIG. 1B, the WTRU 102 may include a processor 118, a transceiver120, a transmit/receive element 122, a speaker/microphone 124, akeypad 126, a display/touchpad 128, non-removable memory 106,removable memory 132, a power source 134, a global positioningsystem (GPS) chipset 136, and other peripherals 138. It will beappreciated that the WTRU 102 may include any sub-combination ofthe foregoing elements while remaining consistent with anembodiment.

[0050] The processor 118 may be a general purpose processor, aspecial purpose processor, a conventional processor, a digitalsignal processor (DSP), a plurality of microprocessors, one or moremicroprocessors in association with a DSP core, a controller, amicrocontroller, Application Specific Integrated Circuits (ASICs),Field Programmable Gate Array (FPGAs) circuits, any other type ofintegrated circuit (IC), a state machine, and the like. Theprocessor 118 may perform signal coding, data processing, powercontrol, input/output processing, and/or any other functionalitythat enables the WTRU 102 to operate in a wireless environment. Theprocessor 118 may be coupled to the transceiver 120, which may becoupled to the transmit/receive element 122. While FIG. 1B depictsthe processor 118 and the transceiver 120 as separate components,it will be appreciated that the processor 118 and the transceiver120 may be integrated together in an electronic package orchip.

[0051] The transmit/receive element 122 may be configured totransmit signals to, or receive signals from, a base station (e.g.,the base station 114a) over the air interface 116. For example, inone embodiment, the transmit/receive element 122 may be an antennaconfigured to transmit and/or receive RF signals. In anotherembodiment, the transmit/receive element 122 may be anemitter/detector configured to transmit and/or receive IR, UV, orvisible light signals, for example. In yet another embodiment, thetransmit/receive element 122 may be configured to transmit andreceive both RF and light signals. It will be appreciated that thetransmit/receive element 122 may be configured to transmit and/orreceive any combination of wireless signals.

[0052] In addition, although the transmit/receive element 122 isdepicted in FIG. 1B as a single element, the WTRU 102 may includeany number of transmit/receive elements 122. More specifically, theWTRU 102 may employ MIMO technology. Thus, in one embodiment, theWTRU 102 may include two or more transmit/receive elements 122(e.g., multiple antennas) for transmitting and receiving wirelesssignals over the air interface 116.

[0053] The transceiver 120 may be configured to modulate thesignals that are to be transmitted by the transmit/receive element122 and to demodulate the signals that are received by thetransmit/receive element 122. As noted above, the WTRU 102 may havemulti-mode capabilities. Thus, the transceiver 120 may includemultiple transceivers for enabling the WTRU 102 to communicate viamultiple RATs, such as UTRA and IEEE 802.11, for example.

[0054] The processor 118 of the WTRU 102 may be coupled to, and mayreceive user input data from, the speaker/microphone 124, thekeypad 126, and/or the display/touchpad 128 (e.g., a liquid crystaldisplay (LCD) display unit or organic light-emitting diode (OLED)display unit). The processor 118 may also output user data to thespeaker/microphone 124, the keypad 126, and/or the display/touchpad128. In addition, the processor 118 may access information from,and store data in, any type of suitable memory, such as thenon-removable memory 106 and/or the removable memory 132. Thenon-removable memory 106 may include random-access memory (RAM),read-only memory (ROM), a hard disk, or any other type of memorystorage device. The removable memory 132 may include a subscriberidentity module (SIM) card, a memory stick, a secure digital (SD)memory card, and the like. In other embodiments, the processor 118may access information from, and store data in, memory that is notphysically located on the WTRU 102, such as on a server or a homecomputer (not shown).

[0055] The processor 118 may receive power from the power source134, and may be configured to distribute and/or control the powerto the other components in the WTRU 102. The power source 134 maybe any suitable device for powering the WTRU 102. For example, thepower source 134 may include one or more dry cell batteries (e.g.,nickel-cadmium (NiCd), nickel-zinc (NiZn), nickel metal hydride(NiMH), lithium-ion (Li-ion), etc.), solar cells, fuel cells, andthe like.

[0056] The processor 118 may also be coupled to the GPS chipset136, which may be configured to provide location information (e.g.,longitude and latitude) regarding the current location of the WTRU102. In addition to, or in lieu of, the information from the GPSchipset 136, the WTRU 102 may receive location information over theair interface 116 from a base station (e.g., base stations 114a,114b) and/or determine its location based on the timing of thesignals being received from two or more nearby base stations. Itwill be appreciated that the WTRU 102 may acquire locationinformation by way of any suitable location-determination methodwhile remaining consistent with an embodiment.

[0057] The processor 118 may further be coupled to otherperipherals 138, which may include one or more software and/orhardware modules that provide additional features, functionalityand/or wired or wireless connectivity. For example, the peripherals138 may include an accelerometer, an e-compass, a satellitetransceiver, a digital camera (for photographs or video), auniversal serial bus (USB) port, a vibration device, a televisiontransceiver, a hands free headset, a Bluetooth.RTM. module, afrequency modulated (FM) radio unit, a digital music player, amedia player, a video game player module, an Internet browser, andthe like.

[0058] FIG. 1C is a system diagram of the RAN 104 and the corenetwork 106 according to an embodiment. As noted above, the RAN 104may employ an E-UTRA radio technology to communicate with the WTRUs102a, 102b, 102c over the air interface 116. The RAN 104 may alsobe in communication with the core network 106.

[0059] The RAN 104 may include eNode-Bs 140a, 140b, 140c, though itwill be appreciated that the RAN 104 may include any number ofeNode-Bs while remaining consistent with an embodiment. TheeNode-Bs 140a, 140b, 140c may each include one or more transceiversfor communicating with the WTRUs 102a, 102b, 102c over the airinterface 116. In one embodiment, the eNode-Bs 140a, 140b, 140c mayimplement MIMO technology. Thus, the eNode-B 140a, for example, mayuse multiple antennas to transmit wireless signals to, and receivewireless signals from, the WTRU 102a.

[0060] Each of the eNode-Bs 140a, 140b, 140c may be associated witha particular cell (not shown) and may be configured to handle radioresource management decisions, handover decisions, scheduling ofusers in the uplink and/or downlink, and the like. As shown in FIG.1C, the eNode-Bs 140a, 140b, 140c may communicate with one anotherover an X2 interface.

[0061] The core network 106 shown in FIG. 1C may include a mobilitymanagement gateway (MME) 142, a serving gateway 144, and a packetdata network (PDN) gateway 146. While each of the foregoingelements are depicted as part of the core network 106, it will beappreciated that any one of these elements may be owned and/oroperated by an entity other than the core network operator.

[0062] The MME 142 may be connected to each of the eNode-Bs 142a,142b, 142c in the RAN 104 via an S1 interface and may serve as acontrol node. For example, the MME 142 may be responsible forauthenticating users of the WTRUs 102a, 102b, 102c, beareractivation/deactivation, selecting a particular serving gatewayduring an initial attach of the WTRUs 102a, 102b, 102c, and thelike. The MME 142 may also provide a control plane function forswitching between the RAN 104 and other RANs (not shown) thatemploy other radio technologies, such as GSM or WCDMA.

[0063] The serving gateway 144 may be connected to each of theeNode Bs 140a, 140b, 140c in the RAN 104 via the S1 interface. Theserving gateway 144 may generally route and forward user datapackets to/from the WTRUs 102a, 102b, 102c. The serving gateway 144may also perform other functions, such as anchoring user planesduring inter-eNode B handovers, triggering paging when downlinkdata is available for the WTRUs 102a, 102b, 102c, managing andstoring contexts of the WTRUs 102a, 102b, 102c, and the like.

[0064] The serving gateway 144 may also be connected to the PDNgateway 146, which may provide the WTRUs 102a, 102b, 102c withaccess to packet-switched networks, such as the Internet 110, tofacilitate communications between the WTRUs 102a, 102b, 102c andIP-enabled devices. An access router (AR) 150 of a wireless localarea network (WLAN) 155 may be in communication with the Internet110. The AR 150 may facilitate communications between APs 160a,160b, and 160c. The APs 160a, 160b, and 160c may be incommunication with STAs 170a, 170b, and 170c.

[0065] The core network 106 may facilitate communications withother networks. For example, the core network 106 may provide theWTRUs 102a, 102b, 102c with access to circuit-switched networks,such as the PSTN 108, to facilitate communications between theWTRUs 102a, 102b, 102c and traditional land-line communicationsdevices. For example, the core network 106 may include, or maycommunicate with, an IP gateway (e.g., an IP multimedia subsystem(IMS) server) that serves as an interface between the core network106 and the PSTN 108. In addition, the core network 106 may providethe WTRUs 102a, 102b, 102c with access to the networks 112, whichmay include other wired or wireless networks that are owned and/oroperated by other service providers.

[0066] FIG. 1D is a diagram of an example use of beamforming in aWLAN 185. The WLAN 185 may include an AP 190 and an STA 192 forminga BSS. FIG. 1E is a diagram of an example use of beamforming usingspatial diversity or multipath diversity in the WLAN 185. The WLAN185 may include an AP 190 and an STA 192 forming a BSS. A WLAN in aInfrastructure Basic Service Set (BSS) mode has an Access Point(AP) 190 for the BSS and one or more stations (STAs) 192 associatedwith the AP. The AP 190 may have an access, or interface, to aDistribution System (DS) 195, or another type of wired/wirelessnetwork that carries traffic in and out of the BSS. Traffic to STAsthat originates from outside the BSS may arrive through the AP tobe delivered to the STAs. Traffic originating from STAs todestinations outside the BSS may be sent to the AP to be deliveredto the respective destinations. Traffic between STAs within the BSSmay also be sent through the AP where the source STA may sendtraffic to the AP and the AP may deliver the traffic to thedestination STA. Such traffic between STAs within a BSS may bepeer-to-peer traffic. Such peer-to-peer traffic may also be sentdirectly between the source and destination STAs with a direct linksetup (DLS) using an 802.11e DLS or an 802.11z tunneled DLS (TDLS).A WLAN using an Independent BSS (IBSS) mode has no AP, and/or STAs,communicating directly with each other. This mode of communicationmay be referred to as an "ad-hoc" mode of communication.

[0067] As used herein an STA 192 may include, but is not limitedto, a WTRU 102, an AP, or a communication device. Using the 802.11infrastructure mode of operation, the AP 190 may transmit a beaconon a fixed channel, usually the primary channel. This channel maybe 20 MHz wide, and may be the operating channel of the BSS. Thischannel may also be used by the STAs to establish a connection withthe AP. The fundamental channel access mechanism in an 802.11system may be Carrier Sense Multiple Access with CollisionAvoidance (CSMA/CA). In this mode of operation, every STA,including the AP, may sense the primary channel. If the channel isdetected to be busy, the STA may back off. Hence only one STA maytransmit at any given time in a given BSS.

[0068] In an 802.11n example, High Throughput (HT) STAs may alsouse a 40 MHz wide channel for communication. This 40 MHz widechannel may be achieved by combining the primary 20 MHz channel,with an adjacent 20 MHz channel to form a 40 MHz wide contiguouschannel. 802.11n may operate on the 2.4 GHz, and 5 GHz ISMbands.

[0069] In an 802.11ac example, Very High Throughput (VHT) STAs maysupport 20 MHz, 40 MHz, 80 MHz, and 160 MHz wide channels. The 40MHz, and 80 MHz, channels may be formed by combining contiguous 20MHz channels similar to 802.11n described above. A 160 MHz channelmay be formed either by combining 8 contiguous 20 MHz channels, orby combining two non-contiguous 80 MHz channels, this may also bereferred to as an 80+80 configuration. For the 80+80 configuration,the data, after channel encoding, is passed through a segmentparser that may divide it into two streams. IFFT and time domainprocessing are done on each stream separately. The streams may thenbe mapped on two channels, and the data may be transmitted. At thereceiver, this process may be reversed, and the combined data maybe sent to the MAC. 802.11ac may operate only on the 5 GHz ISMband, and consequently may not be backward compatible with 802.11nmodes of operation in the 2.4 GHz ISM band. For the examplesdescribed herein, any combination of channels may be used, andshould not be limited to contiguous and non-contiguouschannels.

[0070] Sub 1 GHz modes of operation may be supported by 802.11af,and 802.11ah. For these specifications the channel operatingbandwidths may be reduced relative to those used in 802.11n, and802.11ac. 802.11af may support 5 MHz, 10 MHz and 20 MHz bandwidthsin the TV White Space (TVWS) spectrum, and 802.11ah may support 1MHz, 2 MHz, 4 MHz, 8 MHz, and 16 MHz bandwidths using a non-TVWSspectrum. A possible use case for 802.11ah may support Meter TypeControl (MTC) devices in a macro coverage area. MTC devices mayhave limited capabilities including only support for limitedbandwidths, but also may include a requirement for a very longbattery life. 802.11ah may also be used for macro coverage assupport for cellular offload to WiFi.

[0071] In 802.11ad, wide bandwidth spectrum at 60 GHz may beavailable, thus enabling very high throughput operation. 802.11admay support up to 2 GHz operating bandwidths and the data rate mayreach up to 6 Gbps. Since the propagation loss at 60 GHz may bemore significant than at the 2.4 GHz, and 5 GHz bands, beamformingmay be adopted in 802.11ad as a means to extend the coverage range.To support the receiver requirements for this band, the 802.11acMAC layer may be modified in several areas. An importantmodification for the 802.11ad MAC layer may include procedures thatallow channel estimation and training. These procedures may includeomni, and beamformed modes of operation which do not exist in802.11ac.

[0072] WLAN systems that support multiple channels and channelwidths, such as 802.11n, 802.11ac, 802.11af, and 802.11ah, mayinclude a channel designated as the primary channel. The primarychannel may, but not necessarily, have a bandwidth equal to thelargest common operating bandwidth supported by all STAs in theBSS. The bandwidth of the primary channel may therefore be limitedby the STA, of all the STAs operating in a BSS, which supports orenables the use of the smallest bandwidth operating mode. In theexample of 802.11ah, the primary channel may be 1 MHz wide if thereare STAs, for example, MTC type devices that only support a 1 MHzmode, even if the AP, and other STAs in the BSS, may support a 2MHz, 4 MHz, 8 MHz, 16 MHz, or other channel bandwidth operatingmodes. All carrier sensing and NAV settings may depend on thestatus of the primary channel. For example, if the primary channelis busy due to a STA supporting only a 1 MHz operating mode, thenthe entire available frequency bands may be considered busy eventhough a majority of the frequency bands remain idle andavailable.

[0073] In the United States, for example, the available frequencybands that may be used by 802.11ah are from 902 MHz to 928 MHz. InKorea, for example, the available frequency bands may be from 917.5MHz to 923.5 MHz, and in Japan, it may be from 916.5 MHz to 927.5MHz. The total bandwidth available for 802.11ah may be 6 MHz to 26MHz depending on the country.

[0074] One of the challenges for wireless communication over 60 GHzmay include significant propagation loss due to the high frequency.As the wavelength decreases, the free space propagation loss mayincrease. To address the range limitation due to this propagationloss, 802.11ad may use beamforming to increase the EffectiveRadiated Power (ERP) of the transmissions. Since the wavelength issmall, it may be possible to use a large antenna array to get avery high beamformed antenna gain. The beam in 802.11ad may beelectronically steered to a particular STA, or group of STAs,during association with the STAs.

[0075] In order to support beamforming, the 802.11ad PHY and MACspecifications may be modified to support directionaltransmissions, and millimeter wave (mmW) antenna trainingprocedures. A comprehensive beamforming training protocol may bedefined in 802.11ad. The beamforming training protocol may includetwo components, for example a sector level sweep (SLS) procedureand a beam refinement protocol (BRP) procedure. The SLS proceduremay be used for transmit beamforming training. The BRP proceduremay enable receive beamforming training and iterative refinement ofboth the transmit and receive beams.

[0076] In order to reduce implementation complexity, 802.11ad maysupport beam switching at both the AP and the STA. The beamswitching at both the AP and the STA may be in contrast to moreadvanced multi-antenna schemes, and may assume that a single RFfront-end is available at both ends. Problems related to linkrobustness and spectral efficiency may be of importance forenabling 802.11ad+ to address the current trend for a Carrier GradeWiFi service. A Carrier Grade WiFi service may be referred to as 5GCarrier Grade WiFi, may provide high air interference efficiencyfor multiple users, and a stable "cellular-like" quality. A 5GCarrier Grade WiFi system may support robust and dynamicdeployments, for example dense deployments and flash crowds.

[0077] Although 802.11ad may address the need for a very high peakthroughput, limitations due to the propagation environment are notadequately addressed in 802.11ad.

[0078] For mmW communications it may be necessary to handle apropagation loss due in part to the high free space propagationloss which may occur at mmW frequencies. For example, blockage oftransmissions by the human body may attenuate a signal by 15 to 25dB for hundreds of milliseconds.

[0079] While propagation loss of walls and other indoor obstaclesmay prevent the propagation of mmW through them, indoor Line ofSite (LOS) propagation may occur indoors. LOS propagation may occurindoors either due to a direct line of sight transmission, or dueto reflections off walls and other obstacles. It should be notedthat non-negligible propagation loss at mmW frequencies may occurdue to walls and other environmental factors. Beam switching may beused in 802.11ad to utilize signal diversity due to thesereflections.

[0080] Some mmW communications may utilize a single beam forcommunications, for example, in 802.11ad. MIMO techniques, such asspatial multiplexing, may be employed to improve the spectralefficiency of the system, however it may be difficult to use thesetechniques in mmW systems due to the need for multiple symbolgeneration. Methods that improve the spectral efficiency of mmWsystems, such as methods that enable the use of spatialmultiplexing, may be needed in future mmW specifications which forexample may be based on 802.11ad, and/or in mmW systems ingeneral.

[0081] Beamforming training protocols, such as SLS and BRP, may beused to perform transmit/receive beamforming training and iterativebeam refinement training. The beamforming training overhead andlatency of these procedures, however, may be significant. Forexample, with the transmit beamforming training defined in SLS, thetransmitter may need to transmit multiple sector sweep (SSW) framesthat may be modulated by different beamforming sectors. Each devicemay have up to 64 different beam sectors. Each SSW frame mayinclude a full PLCP header, which may include a preamble, one ormore header blocks, and a MAC frame. In order to fully exploit thebeamforming gain, transmit and receive training procedures at bothpeer devices may be required, and an iterative beam refinement mayalso be needed. These procedures may represent a significantoverhead and hence methods that reduce this overhead may be neededin part to allow for a better user experience in mmW systems.

[0082] Referring back to FIGS. 1D and 1E, multipath propagation maybe common in indoor communications links. Beam switching basedbeamforming algorithms utilized in the current 802.11adspecification may attempt to point the beam to the strongest path.As shown in FIG. 1D, a LOS path and a strong reflection path mayexist between AP 190 and STA 192. After the beamforming trainingprocedure, the beam with the best channel gain may be selected.This beam may be formed towards the strongest path among multiplepropagation paths. However, human blockage may introduce an average20 dB loss for 230 ms, which may prevent the 60 GHz radio toprovide multi-Gigabit/sec data transmissions. It is thereforehighly likely that the beamformed link may be dropped, and hencethe transmitted packet during this period would be lost. Moreover,due to the loss of the packet, the system may have to repeat thebeamforming training and then retransmit the packet over apotentially new beam.

[0083] Example beamforming schemes may exploit the spatialdiversity such that the transmission is not dependent on only thestrongest path. As shown in FIG. 1E, with two strong propagationpaths, for example, a LOS path and a strong reflection path, anumber of solutions may be possible including fast beam switching,wider beam and multi-beam methods.

[0084] Spatial diversity may be achieved with fast beam switching.In order to accomplish fast beam switching when the channelcondition changes, it may be necessary that both AP and STA have anavailable list of weight vectors/beam identifiers. Example methodsfor obtaining a list of weight vectors/beam identifiers aredisclosed below. There may be two possibilities for fast-beamswitching including in-band signaling and out-of-band signaling.In-band beam switching may be used in some examples.

[0085] In an in-band beam switching example, the AP may becommunicating with the STA using a beam set (Tx and Rx beams) B1.The AP and STA may have prioritized the beam sets according to thereceived SNR, or SINR, during the SLS and BRP phases. The STA maymonitor one or more of the received SNR, SINR, Bit-Error-Rate (BER)or Packet Error Rate (PER), acknowledgement (ACK) statistics or acombination of these parameters. If, at the end of a packetreception, the STA determines that the channel quality isdeteriorating, it may append a message to the ACK packet requestingthe AP to switch to the next best beam set, for example B2, for thenext transmitted packet. The assumption here is that the channelcondition may be deteriorating, but not to the point that the datapacket cannot be decoded correctly and hence an ACK may besent.

[0086] In another example, alternative beam retransmission methodsmay be used. In 802.11 systems, and mmW systems in particular, ifno ACK is received, the data packet may be retransmitted. Thisretransmission may use the same beam that was used in the priortransmission. An example for an alternative procedure may be thatif the AP does not receive an ACK, the AP may retransmit the datapacket using the beam set B2 instead of the beam set B1. Sincethese beam sets may have been defined prior to this procedure,possibly during association of the STA with the AP, the procedureat the STA may use the corresponding receive beam set B2 forreception of the retransmission from the AP. In an exampleprocedure, the AP and the STA may define an association of indicesto beam sets, and subsequently use the beam set indices foridentification of beam sets in the aforementioned procedures.

[0087] In an alternative, or additional example, the AP may cyclethe data packet through the N best beam-sets. The STA may thenperform a procedure wherein it receives N packets from the AP, andperform maximum-ratio-combining, selection combining, or a similarreceive algorithm, on these packets. An ACK may then be determinedand sent to the AP, for example, after all N transmissions haveoccurred, or as soon as the packet has been successfully receivedand decoded.

[0088] In an alternative, or additional example, the AP may alsotransmit the data packet through all of, or a subset of, the Nbeam-sets simultaneously. The remainder of the procedure describedin the previous paragraph may then follow in a similar way.

[0089] A beam set may include the identification of a primary beamwithin the beam set. The primary beam may be used by transmissionprocedures at the AP, STA, or both, as the beam to be used forinitial attempts at wireless communication. Alternatively, theprimary beam may be used exclusively for transmission of control orscheduling information. More than one primary beam may be used formore than one STA wherein each primary beam may be associated witha particular STA.

[0090] In an alternative, or in addition to, the AP may also cyclethrough different modes of MIMO operation for each beam set, priorto proceeding to the next beam set. For example, if the AP does notreceive an ACK on beam set B1 it may select a more robust form ofoperation such as Space Time Block Coding (STBC), Space FrequencyBlock Coding (SFBC), or Cyclic Delay Diversity (CDD), beforeproceeding to transmit on the remaining beam sets B2 throughBN.

[0091] In some examples, a first and second best beam transmissionmethod may be used. For example, the STA may determine the two bestbeams, B1, B2, using a procedure similar to the above, andrecommend these beam indices to the AP. In this example, for theremainder of the communication interval, the AP and the STA mayassume that either, or both, beam indices may be used for wirelesscommunication.

[0092] The AP may then determine to transmit on either beam duringa particular transmission time interval (TTI) based on one or morecriteria determined by the AP. Example criteria may include one ormore of the received SNR, or SINR, Bit-Error-Rate (BER), or PacketError Rate (PER), acknowledgement (ACK) statistics, or acombination of these criteria. The STA may respond to a messagefrom the AP with an indication of the reception quality, or similarmetric, for the beam that it used to receive the message from theAP. This indication or metric may be indicated in the response bythe beam index.

[0093] Alternatively, if the STA does not provide an indication ofthe reception quality in its ACK and/or any other packet to the AP,the AP may determine that the reception quality was acceptable forone or more associated beams. If the STA indicates a poor receptionquality to the AP for B1, it may assume that the next transmissionfrom the AP will use the second best beam B2.

[0094] During communication with the STA, the AP may store thepacket reception quality for each transmission, on each beam, in amemory, while using either beam. If a particular beam in the pairbecomes unusable for further communication, the AP may identify anew first, or a second best beam for communication with the STA,while at the same time continuing communication on the remainingbeam of the original beam pair. Other combinations of theprocedures described in this example may be possible. The aboveexample is not limited to a pair of beams and may be extended tosupport any number of beams greater than one that the system maysimultaneously support.

[0095] Some examples may use out of band beam switching. Forexample, the AP and STA may both have multi-band capability. Inthis example, the AP and the STA may communicate over either a 2.4GHz or a 5 GHz link in addition to a directional 60 GHz link. TheAP and the STA may use one or more of the sub-6 GHz links as analternate link to signal to each other that the beam set needs toswitch to the next best beam set at the beginning of the next datapacket. This example procedure may allow for a fast beam switch tooccur even if the packet currently being transmitted is not decodedcorrectly.

[0096] Some examples may implement spatial diversity using a singlebeam. For example, it may be possible that only one RF chain isavailable at both transmitter and receiver, such that only one datastream may be transmitted and received at the same time. This RFchain arrangement may be used in mmW systems. With one RF chain,devices may form one beam and transmit the data stream toward thedirection specified by the beam. In this example, the devices mayforming a beam pointing to a propagation path with the strongestchannel gain. Multipath wireless channels, however, may introducefrequency selectivity. A beamforming weight may benefit somefrequency tones, however, it may have a detrimental effect for oneor more of the other set of frequency tones. Accordingly, there maybe no guarantee that the weight pointing to the strongestpropagation path will introduce the maximum beamforming gain forthe entire frequency channel. Moreover, pointing in one beamdirection may increase the system sensitivity to small changes inthe multipath environment and may fail to provide robustcommunication.

[0097] FIG. 2 is a diagram of an example method 200 using two STAsto perform multi-path beamforming. The method may be a multi-stageiterative beamforming method as discussed below, and may includegrouping two or more antennas. For example, FIG. 2 shows a firstiteration 201 of the method and a second iteration 202 of themethod.

[0098] The following articles disclose methods for beamforming. P.Xia, S. K. Yong, J. Oh and C. Ngo, "A practical SDMA protocol for60 GHz millimeter wave communications", Asilomar, 2008; and, P.Xia, S. K. Yong, J. Oh and C. Ngo, "Mulit-stage iterative antennatraining for millimeter wave communications", Globecom, 2008, theentire contents of both are incorporated herein by reference.

[0099] Referring to FIG. 2, STA.sub.1 205 and STA.sub.2 210 arerespectively shown on two time axes 215 and 220. In this example,STA.sub.1 205 may transmit one or more training beamformingweights, and STA.sub.2 210 may receive training beamforming weights230, 240, and transmit beamforming weights 235 to STA.sub.1 205.STA.sub.1 205 and STA.sub.2 210 may be, for example, WTRU,stations, electronic communication devices, or access points. Theexample in FIG. 2 shows STA.sub.1 205 transmitting 230, 240 to onlySTA.sub.2 210, however there may be more than one STA, which arenot shown. STA.sub.1 205 may be an AP or a non-AP STA. STA.sub.2210 may be an AP or a non-AP STA.

[0100] In each iteration of the method 201, 202, the AP, hereSTA.sub.1 205, may transmit one or more training sequences overmultiple time slots 230 and sweep the transmit beamforming weights.For example, STA.sub.1 205 may be an AP or a non-AP STA. Note thatonly two iterations of the method 201, 202 are illustrated, but themethod may have more than two iterations of the method 201, 202.STA.sub.2 210 may calculate the best transmit beamforming weightvector according to an estimate of the received channel state. Notethat the calculated transmit beamforming weight may not be one ofthe weights the transmitter STA.sub.1 205 originally utilized. Themethod at STA.sub.2 210 may then feedback this beamforming weightvector 235, or the estimated channel state vector, to the AP, hereSTA.sub.1 205. The AP, here STA.sub.1 205, may, or may not, updatethe transmit beamforming weight it utilizes for a subsequenttransmission to STA.sub.2 210. The method may continue until packettransmissions are completed for the associated STAs, STA.sub.2210.

[0101] The AP, here STA.sub.1 205, may group the antenna array intomultiple sub-groups to point the beamforming weight to multiplepaths. For example, if there are 36 antenna units and only the twostrongest paths are used, then each sub-group may have 18 antennaunits. Alternatively, if more antenna gain is anticipated from thestrongest path, more antenna units may be assigned to the strongestpath. The AP, here STA.sub.1 205, may assign other antennasub-group partitions depending on the requirements of the system.For example, more than two strongest paths may be used. The methodin this example may steer the antenna array in the first sub-groupto the strongest path, while the second sub-group may be steered tothe second strongest path, and so on. This antenna group partitionprocedure may be performed by the AP, here STA.sub.1 205, orSTA.sub.2 210, or both.

[0102] In another example, antenna group based multi-pathbeamforming may be performed. In this example, STA.sub.1 205 mayhave Nt transmit antennas and STA.sub.2 210 may have Nr receiveantennas, and only the two strongest paths may be considered.

[0103] The transmitter, for example STA.sub.1 205, may transmit Ntsequences 230. The Nt sequences 230 may be modulated using Ntorthogonal beamforming vectors. STA.sub.1 205 may include aprecoder, for example an identity precoder, and may be configuredto transmit the first sequence using the first antenna, andtransmit the second sequence using the second antenna, and so on.Other orthogonal precoding matrices may be utilized by STA.sub.1205.

[0104] The receiver, for example STA.sub.2 210, may receive the Ntsequences using the receive beamforming vector calculated in thelast iteration, W.sub.i-1.sup.r. In some examples, W.sub.i-1.sup.rmay be set to an initial value. STA2 210 may be configured todetermine that the first n time slots correspond to transmitantenna 1 to antenna n, which may correspond to antenna group 1.STA2 210 may utilize the training sequence transmitted in each timeslot to estimate the strongest path of the propagation channel, andmay denote the strongest channel path as H1. The received signalfrom the first n time slots may be expressed asy.sub.i.sup.11=W.sub.i-1.sup.r(1:n)Hs+N, where y may be thereceived symbol, s may be the sent symbol, N may be the additiveGaussian noise having a variance, and H may be the channel matrixbetween the transmitter STA1 205 and the receiver STA2 210. Thereceiver STA2 210 may use the correlation property of the trainingsequence to estimate the channel corresponding to the strongestpropagation path, H.sub.1.sup.1. Thus, the receiver STA2 210 maydetermine the best transmit beamforming weight corresponding toantenna group 1 and the strongest path, and may be represented as(H.sub.1.sup.1)'. The size of the beamforming weight may ben.times.1.

[0105] The receiver, STA2 210, may be configured to determine thattime slots n+1 to Nt correspond to transmit antenna n+1 to antennaNt, and may correspond to antenna group 2 of STA1 205. Thereceiver, STA2 210, may utilize the training sequence transmittedin each time slot to estimate the second strongest path of thepropagation channel, and may be denoted as H2. The received signal245 from the Nt-(n+1)+1 time slots may be expressed asy.sub.i.sup.12=W.sub.i-1.sup.r(n+1:Nt)Hs+N Since the signal may betransmitted using a sequence with a zero auto correlation (ZAC)property, STA2 210 may use a Rake receiver like method, where thestrongest path may be removed, and the channel of the secondstrongest path may be determined. The second strongest path may berepresented by H.sub.2.sup.1. Thus the best transmit beamformingweight corresponding to antenna group 2 and the second strongestpath may be (H.sub.2.sup.1)'. The size of the transmit beamformingweight may be (Nt-n).times.1.

[0106] The updated transmit beamforming weight for iteration i maybe expressed as W.sub.i.sup.t=[H.sub.1.sup.1,H.sub.2.sup.1]'. STA2210 may transmit W.sub.i.sup.t back to STA1 205 at 235. STA1 205may use the received W.sub.i.sup.t to transmit 240 another set ofbeamforming training frames. STA1 205 may transmit Nr repetition oftraining frames, where Nr may be the number of antenna(s) at STA2205. STA2 205 may be configured to use this set of training framesto update the received beamforming weight. STA2 210 may use Nrantennas to receive the training frames sequentially. STA2 210 mayalso use other orthogonal beamforming weights to receive the Nrtraining frames. In this example, at STA2 210, antenna 1 to m maybelong to the first antenna group, and may be used to point to thestrongest propagation path, while antenna m+1 to Nr may belong toantenna group 2, and may correspond to the second strongestpropagation path.

[0107] In one example, STA1 205 may use a mixed mode to transmit Nrrepetitions of training frames with weight W.sub.i.sup.t. Thisexample mixed mode method is shown in FIG. 2 as 202. The receivedsignal through the Nr time slot may be expressed asy.sub.i=HW.sub.i.sup.ts+n. The received signal from antenna group 1may be y.sub.i.sup.21=y.sub.i(1:m) STA2 210 may use the correlationproperty of the training sequence to estimate the channel with thestrongest propagation path, H.sub.1.sup.2. The received signal fromantenna group 2 may be y.sub.i.sup.22=y.sub.i(m+1:Nr) and STA2 210may use correlation detection to remove the strongest path. STA2210 may accordingly determine the estimated channel for the secondstrongest path, H.sub.2.sup.2. STA2 210 may update the receivebeamforming weight, which may be represented as,W.sub.i.sup.r=[H.sub.1.sup.2,H.sub.2.sup.2]'. In FIG. 2, thereception of the signals is shown in dashed lines as an exampleillustration of the receiver operation while receiving apacket.

[0108] Alternatively, or in addition to, STA1 205 may use asequential mode. STA1 205 may transmit m repetitions of trainingframes with antenna group 1, i.e.,W.sub.1.sup.t1=[H.sub.1.sup.1,0]'. During the m training time slot,the receiver, for example STA2 210, may utilize antenna group 1 toreceive the training frames. For example, STA2 210 may utilizeantenna 1 to receive the first training frame, and antenna m toreceive the mth training frame. The received signal from these mframes may be expressed as y.sub.i.sup.21=HW.sub.i.sup.t1s+n. STA2210 may estimate the channel corresponding to the strongestpropagation path H.sub.1.sup.2. STA1 205 may transmit Nr-mrepetitions of training frames with antenna group 2 only, forexample, W.sub.1.sup.t2=[0,H.sub.2.sup.1]'. STA2 210 may utilizeits antenna group 2 to receive the training frames. The receivedsignal may be y.sub.i.sup.22=HW.sub.i.sup.t2=HW.sub.i.sup.t2s+n.STA2 210 may use a correlation method to determine the channelcorresponding to the second strongest path, H.sub.2.sup.2. STA2 210may update the receive beamforming weightW.sub.i.sup.r=[H.sub.1.sup.2,H.sub.2.sup.2]'. In some examples,feedback channels corresponding to the strongest paths may be useddirectly by STA2 210 or STA1 205. Note that this mode is notillustrated in FIG. 2.

[0109] STA2 210 may sendW.sub.i.sup.r=[H.sub.1.sup.2,H.sub.2.sup.2]' to STA1 205 (notillustrated). In addition, the method may repeat for a number oftimes up to a threshold or until the method converges which may bedetermined by comparing a next value of the beamforming weightswith a previous value of the beamforming weights and determining ifthe difference is less than a threshold value.

[0110] The initial beamforming weights for STA1 205 and STA2 210may be set to initial values prior to beginning the method. STA1205 and STA2 210 may determine the initial values in order toreduce the number iterations needed for the method to converge. Theexample method shown in FIG. 2 may be used for determining the twostrongest paths. However, in other example embodiments, more thantwo paths may be determined by the STA1 205 and STA2 210.

[0111] Existing protocols may be modified to perform multi-pathbeamforming methods. For example, a multi-path beamforming methodmay be used in 802.11 and 802.11ad. For example, the multi-pathbeamforming method may be used in 802.11ad by using a modificationof the beam refinement protocol (BRP) as disclosed in IEEEP802.11ad.TM./D9.0: Part 11, "Wireless LAN Medium Access Control(MAC) and Physical Layer (PHY) specifications," the entire contentsof which are herein incorporated by reference. The multi-pathbeamforming method may be a multi-iteration multi-path beamformingtraining method.

[0112] A beam refinement transaction may be a set of BRP framesthat include one or more beam refinement requests and responses.The multi-path beamforming method may be implemented by modifyingcurrent beamforming refinement protocols.

[0113] FIG. 3 is a diagram of an example of one iteration of themulti-path beamforming method 300 using BRP transactions. Thebeamforming initiator, STA1 305, may transmit a BRP frame 315 thatindicates that the BRP frame 315 is a Transmit BRP Request frame.This indication may be performed by setting a field, for example,TX-TRN-REQ=1. A BRP frame with TX-TRN-REQ=1 315 may include atransmit training subfield (TRN-T) 320 appended to it. Theresponder, STA2 310 may reply with a Transmit BRP Feedback 325, forexample, by setting TX-TRN-RSP=1. Moreover, STA2 310 may request areceive beamforming training by indicating Receive BRP Requestframe in the same BRP frame, for example, by setting L_RX>0. Inthis example, the Receive BRP Request frame may be piggybacked onthe Transmit BRP Response frame. L_RX may be a signal field used toindicate that the receiver requests a receive BRP training, and thetransmitter may respond with a BRP train response followed by aTRN-R training field. STA1 305 may transmit a BRP frame 330 with aBRP train response, for example, by setting RX-TRN-RSP to 1. A BRPframe that includes an RX-Train-response that equals 1 may includea receive training subfield (TRN-R) 335 appended to it.

[0114] FIG. 4 is a diagram of an example frame format for a BRPpacket 402. The BRP packet 402 may include a short training field(STF) 404, a channel estimation (CE) field 406, a Header field 408and a data field 410. A training field 416 may beappended/prepended to the BRP packet 402 and may include an AGCtraining field 412 and a receiver/transmitter training subfield(TRN-R/T) 414. A BRP packet 402 may be transmitted using controlPHY. Prior to the RTN-R/T training, there may be a signalingexchange to aid the training procedure. This is the purpose offield 404, 406, 408, 410. A Packet Type field may be included inthe PHY header, and it may indicate whether a TRN-R or a TRN-Tsubfield 414 is appended to the frame 402. A BRP frame 402 with aTRN-R/T 414 field appended may be referred to as a BRP-RX/TX packet402. In a BRP-TX packet 402, the transmitter may change the TXbeamforming weight configuration at the beginning of each AGCsubfield 412. The set of beamforming weights used for the AGCsubfields 412 may be the same as that used for the TRN-T subfield414. In a BRP-RX packet 402, the transmitter may use the sametransmit beamforming weight as in the preamble and data fields ofthe transmission data packet. The BRP frame may be an Action No ACKframe.

[0115] FIG. 5 is a diagram of an example format of a BRP frameAction field. The BRP frame Action field 500 may include a Categoryfield 510, an Unprotected DMG Action field 520, a Dialog Tokenfield 530, a BRP Request field 540, a DMG Beam Refinement element550, and one or more Channel Measurement Feedback elements 5601 . .. 560n.

[0116] An 802.11ad beam refinement protocol may be modified asfollows in order to accommodate a multi-path beamforming algorithm.For example, an initiator may determine the capabilities of theresponder prior to initiating beamforming training with theresponder. The multi-path multi-stage iterative beamformingtraining capability may be indicated in a DMG Capabilities element.A DMG Capabilities element may be present in Association Request,Association Response, Reassociation Request, ReassociationResponse, Probe Request and Probe Response frames and may bepresent in DMG Beacon and Information request and response frames.A DMG Antenna Array Support field may include one or more bits thatindicate that the STA is capable of forming sub-antenna groups andcapable of performing the multi-path multi-stage beamformingtraining method.

[0117] Partitioning of antenna sub-groups at both initiator andresponder may be signaled. Depending on the antenna groupingmethod, the signaling may be different. For example, the antennagrouping may be performed uniformly or non-uniformly.

[0118] In a uniform antenna grouping example, assuming the totalnumber of antennas is even, each antenna sub-group may have thesame number of antenna elements, and hence only the number ofantenna groups is required to indicate the partition of antennasub-groups. For example if there are four antennas, and two groups,the number of groups, in this example two, may be sent back sincethe number of antennas per group will be known. The mapping betweenantenna element indices and sub-group indices may be predeterminedand transmitted explicitly in a field of the BRP frame, forexample, a BRP Request field. If the mapping is explicitlyindicated in the BRP frame, the antenna sub-group index may beassigned to each antenna element.

[0119] In a non-uniform antenna grouping example, each antennasub-group may have a different number of antenna elements. Forexample, the system may assign more antenna elements for thestrongest path, so that the antenna gain from this sub-group may belarger. The mapping between antenna element indices and sub-groupindices may be predetermined and transmitted explicitly in a fieldof the BRP frame, for example, in a BRP Request field.

[0120] Mapping between antenna sub-groups and channel propagationpaths/taps may be predetermined. For example, antenna sub-group 1may always map to the strongest path, and so on. Alternatively, themapping may be defined in the BRP Request field.

[0121] The precoding matrix used by the initiator, for example,STA1 205, in the first part of each iteration may be predeterminedand agreed by both initiator and responder. The first part of eachiteration may be referred to as the transmit beamforming trainingpart. In one example, a set of unitary precoding matrices may bepredetermined. In this example, the initiator and responder maynegotiate which matrix is utilized before performing thebeamforming training. For example, the precoding matrix index maybe predetermined and transmitted in the BRP Request field.

[0122] The number of antennas at both initiator and responder sidemay be signaled. Number of antennas may be signaled, for example,in the PHY header, MAC header or BRP Request field.

[0123] FIG. 6 is a diagram of an example modified channelmeasurement feedback element 600. The channel measurement feedbackelement may include a signal-to-noise ratio (SNR) subfield 610, achannel measurement subfield 620, a tap delay subfield 630, and asector ID order subfield 640. The presence of these subfields maydepend on the values defined in the DMG Beam Refinement element.For example, the channel measurement subfield may be used to feedback up to Ntap channel measurements that correspond to a commonset of relative tap delays defined in the tap delay subfield.Without the presence of the tap delay subfield, for example, theNtaps channel taps may be interpreted as contiguous time samples,separated by Tc, where Tc may be the SC PHY chip time, and may be0.57 ns. In these exemplary multi-path beamforming methods, thechannel measurements of the strongest propagation paths may be sentto STA1. In the example shown in FIG. 2, the strongest path/tap maybe assigned for the first antenna sub-group, and the secondstrongest path/tap may be assigned for the second antenna sub-groupand so on. If the multi-path beamforming method is indicated in thePHY header, MAC header or MAC body, the interpretation of thechannel measurement may be modified when the tap delay subfield isnot present. Therefore, protocols may be modified to accommodateexamples of the multi-path beamforming method disclosed herein.

[0124] Weighted multi-path beamforming training methods may beperformed. For example, a beamforming method for steering the beamtowards multiple propagation paths may be performed. The strongestpropagation paths/taps may be determined by a STA, and one or morebeamforming weights may be determined to point to one or more ofthe propagation paths/taps. The beamforming weight for the kthstrongest propagation path may be represented as Wk, and the finalbeamforming weight may be expressed as

W = k = 1 K .times. .alpha. k .times. W k , ##EQU00001##

where K may be the number of channel propagation paths and.alpha..sub.k may be the weight, with .SIGMA..alpha..sub.k=1.

[0125] Different methods may be used by the STA selecting.alpha..sub.k in one of the following ways. For example,propagation path selection may be based on:

.alpha. k = { 1 .times. .times. k = m 0 .times. .times. k .noteq. m. ##EQU00002##

By this selection, the final beamforming weight vector may equalthe weight vector directed towards the mth propagation path.

[0126] In an 802.11ad example, the channel propagation taps may bemeasured and fed back to the beamforming initiator, which may beSTA1 205. According to the channel measurement of each tap, achannel gain may be estimated by the STA. Channel gain of the kthpropagation path/tap may be represented as .beta..sub.k..alpha..sub.k may be represented as

.alpha. k = .lamda. - 1 .beta. k , ##EQU00003##

so that

.alpha. k = 1 , .lamda. = 1 + 1 .times. / .times. .beta. k K##EQU00004##

may be satisfied. The propagation path with the larger channel gainmay have a larger weight and may be determined by the STA to be thestrongest propagation path.

[0127] A single data stream transmission may be performed withmulti-beam capability devices. For example, multiple RF chains maybe available at the AP. Accordingly, the AP may form multiple beamssimultaneously. In this example, the STA may form only one RFchain. The AP and the STA may be configured to use an Nx1 virtualMIMO channel. The AP and the STA may be configured to use diversitymethods, such as, for example, STBC, SFBC and CDD. These examplemethods may be performed by more than one AP and more than one STA.In addition, a STA may be an AP.

[0128] There may be at least two possible transmission proceduresthat implement RF front-end beamforming with digital domain STBC.One example transmission procedure may use full size beamformingwith STBC. Another example transmission procedure may use partialsize beamforming with STBC.

[0129] FIG. 7 is a diagram of an example AP 700 configured toperform a transmission using full size beamforming with STBC. TheAP 700 may include a coding/modulation unit 702, an STBC encoder706, a plurality of DAC/upconverters 708, 710, and a plurality ofantennas 716. In this example, the coding/modulation unit 702 mayperform modulation and coding and pass the modulation symbols 704to the STBC encoder 706. The STBC encoder 706 may generate two datastreams 718, 720. The two data streams 718, 720 then be processedthrough two RF chains at the plurality of DAC/upconverters 708,710, which may separately perform DAC and up conversion to theoperating frequency band. At the RF front-end, two beamformingweight vectors W1 712 and W2 714 may be generated by AP. Eachweight-vector may be of size Nt.times.1. The first data stream 718may be multiplied with the first weight vector W1 712, and thesecond data stream 720 may be multiplied with the second weightvector W2 514. The two data streams 718, 720 may then be summedtogether and transmitted through the Nt antennas 716. In someembodiments, the AP may be configured with more than two RFchains.

[0130] FIG. 8 is a diagram of an example AP 800 configured toperform a transmission using partial size beamforming with STBC.The AP 800 may include a coding/modulation unit 802, an STBCencoder 806, a plurality of DAC/upconverters 808, 810, and aplurality of sets of antennas 816, 817. In this example, thecoding/modulation unit 802 may perform modulation and coding andpass the modulation symbols 804 to the STBC encoder 806. The STBCencoder 806 may generate two data streams 818, 820. The two datastreams 818, 820 may then be processed through two RF chains at theplurality of DAC/upconverters 808, 810, which may separatelyperform DAC and up conversion to the operating frequency band. Atthe RF front-end, two beamforming weight vectors W1 812 and W2 814may be generated. Each weight vector may be of size Nt/2.times.1.The first data stream 818 may be multiplied with the first weightvector W1 812, and the second data stream 820 may be multipliedwith the second weight vector W2 814. The first data stream 818 maybe transmitted through the first set of Nt/2 antennas 816 and thesecond data stream may be transmitted through the second set ofNt/2 antennas 817. In embodiments, AP may be configured with morethan two RF chains.

[0131] The AP and/or the STA may be configured to send and receivemultiple data streams. For example, the AP and/or the STA may beconfigured with multiple RF chains. In these examples, the AP maybe configured to communicate with multiple STAs simultaneously. TheAP may be configured to distinguish the multiple STAs by spatialdomain beams, thus the method may be referred to as Beam DivisionMultiple Access (BDMA). The AP may need multiple RF chains toperform BDMA. For example, the AP may be configured to use spatialmultiplexing methods for single STA transmission. The AP and STAmay be configured to send more than one data stream at a time,which may increase the spectral efficiency of the system. MultipleRF chains may be needed at both AP and STA side.

[0132] FIG. 9 is a diagram of an example of transceiverarchitecture 900. The AP and/or STA may be configured as follows.The transceiver architecture 900 may include a transmitter side 902and a receiver side 904. The transmitter side 902 may include oneor more coding/modulation units 903, a plurality ofDAC/upconverters 912, 914, a digital controller 917, one or morepower amplifiers (PA)s 920, and a plurality of Nt antennas 925.Multiple data streams may be modulated and coded at baseband, andthen converted from a digital domain to an analog domain throughthe digital controller 917. The streams 908, 910 may be upconvertedto operation frequency band by the DAC/upconverters 912, 914. Twosets of DACs and upconverters are illustrated here, which impliesthat up to two data streams may be supported by the transmitter.Beamforming weights 915, 916 may be applied prior to applying thestreams to the PAs 920. The beamforming weights 915, 916 may beprepared in a digital domain.

[0133] The transmitter 902, may be an AP or STA, an may beconfigured with Nt antennas 925. The Nt antennas 925 may be sharedby two or more RF chains. When the transmitter 902 has two datastreams {s.sub.1,s.sub.2} to transmit, it may generate twobeamforming weights V.sup.1=(V.sub.1.sup.1, V.sub.2.sup.1, . . . ,V.sub.Nt.sup.1).sup.T 916 and V.sup.2=(V.sub.1.sup.2,V.sub.2.sup.2, . . . , V.sub.Nt.sup.2).sup.T 915. The two signalstreams may be combined and transmitted through Nt transmitantennas 925, s=V.sup.1s.sub.1+V.sup.2s.sub.2. The AP and STA maybe similarly configured on the receiver side if multiple RF chainsare presented. The receiver 904 may generate two sets of receivebeamforming weights U.sup.1=(U.sub.1.sup.1, U.sub.2.sup.1, . . . ,U.sub.Nr.sup.1).sup.T 936 and U.sup.2=(U.sub.1.sup.2,U.sub.2.sup.2, . . . , U.sub.Nr.sup.2).sup.T 935 and apply them inan analog domain. The weighted streams may be applied to arespective ADC/downconverter. The downconverted streams 908', 910'may be decoded and demodulated.

[0134] Communication devices, which may be, for example, APs orSTAs, with the transceiver embodiment illustrated in FIG. 9 may bereferred to as Type I. In some embodiments, the AP and/or STA maybe configured with more than two RF chains where FIG. 9 may beextended to accommodate the more than two RF chains.

[0135] FIG. 10 is a diagram of another example transceiverarchitecture 1000. The transceiver architecture 1000 may include atransmitter side 1002 and a receiver side 1004. In this example,the AP and or the STA may be configured as follows. Each RF chainmay have its own set of antennas 1048, 1050, respectively. Theantennas may be deemed to have been partitioned or split intosub-groups based on the number of RF chains. Compared to Type Idevices, in order to achieve the same antenna gain, N antennaelements may be required. Here N may be the number of RF chains.Communication devices, which may be, for example, APs or STA, withthe transceiver architecture illustrated in FIG. 10 may be referredto as Type II. The beamforming weights in FIG. 10, may be tuned asa group by the digital controller logic, as indicated by the dashedlines.

[0136] FIG. 11 is a diagram of an example of beam division multipleaccess (BDMA) architecture 1100. The AP 1102 may be configured totransmit two packets to STA1 1106 and STA2 1104 simultaneously. TheSTAs 1106, 1104 may be configured to share the time-frequencyresource by different RF front-end beams.

[0137] The AP 1102 may be configured to prepare MAC packets forboth STA1 1106 and STA2 1104. The AP 1102 may encode and modulatethe MAC packets and form separate PHY packets and up convert themto 60 GHz through separate RF chains. At the RF front-end, the AP1102 may apply beamforming weight vector W1 to the first datastream and W2 to the second data stream. The AP transmits acombination of the two data streams. In this way, multiple RFchains may share the same set of antennas as shown in FIG. 9.Alternatively, the AP 1102 may be configured to implementation BDMAby dividing, grouping, or partitioning the set of antennas tosub-groups, and each RF chain control may be sent by the AP 1102 onone antenna sub-group as shown in FIG. 10.

[0138] The AP and STA may be configured to perform a beamformingtraining method for BDMA. The AP and/or STA may be configured toperform the beamforming training method sequentially with one ormore STAs in communication with one another. The examplebeamforming training methods may be standardized, for example, andmay be used in 802.11ad. The AP and/or STA may be configured to useorthogonality between training beams.

[0139] FIG. 12 is a diagram of an example beamforming trainingmethod 1200 for BDMA. This example may use a multi-stage iterativebeamforming training algorithm for BDMA. In the exampleillustrated, the AP 1202 has Nt antennas, STA1 1204 has M1 antennasand STA2 1206 has M2 antennas.

[0140] The beamforming training method for BDMA may be performediteratively. The AP 1202 may transmit Nt sequences 1208. The Ntsequences 1208 may be modulated using Nt orthogonal beamformingvectors. The example shown in FIG. 12 may be performed using aprecoder, for example, an identity precoder. In this example, thefirst sequence may be transmitted using a first antenna ("Ant 1")1208, and a second sequence may be transmitted using the secondantenna, and so on until the Nt antenna ("Ant Nt") 1210. The AP1202 may be configured to use other orthogonal precoding matrices,such as, for example, the Walsh Hadamard matrix or FFT matrix.

[0141] STA1 1204 may be configured to utilize the best receivebeamforming vector 1212 calculated through the last iteration,W.sub.i-1.sup.r1, to receive the signals. In FIG. 12, the receptionof the signals is shown in dashed lines as an example illustrationof the receiver operation while receiving a packet. For example,the dashed box W.sub.i-1.sup.r1 may indicate that the STA1 1204should use receive beamforming vector W.sub.i-1.sup.r1 to performreceive beamforming. Since the AP 1202 may have transmitted thesignal through each antenna sequentially, STA1 1204 may receive aNt.times.1 effective MISO channel between AP 1202 and STA1 1204,W.sub.i-1.sup.r1H.sub.1 at the end of the transmission due to theuse of the receive beamformer. Equivalently, the received signalmay be expressed in a matrix format:y.sub.11=W.sub.i-1.sup.r1H.sub.1s+n, where y may be the receivedsignal, s may be the sent signal, n may be Gaussian noise, W may bethe weight used by STA1 1204, and H may be the channel matrixbetween AP 1202 and STA1 1204. Based on the received signal, STA11204 may calculate or determine the best transmit beamformingweight from AP 1202 to STA1 1204 which may be represented as

V i r .times. .times. 1 = y 11 .times. s ' y 11 .times. s ' .##EQU00005##

[0142] Similarly, STA2 1206 may be configured to utilizeW.sub.i-1.sup.r2 to receive the signals, and the received signalmay be expressed as y.sub.12=W.sub.i-1.sup.r2H.sub.2s+n. The besttransmit beamforming weight from AP 1202 to STA2 1206 may berepresented as

V i r .times. 2 = y 1 .times. 2 .times. s ' y 1 .times. 2 .times. s' , ##EQU00006##

which STA2 1206 may be configured to determine.

[0143] The method may continue with STA1 1204 sendingV.sub.i.sup.r1 1222 to the AP 1202. STA2 1206 may sendV.sub.i.sup.r2 1220 to the AP 1202. STA2 1206 may transmit thepacket immediately after the transmission of STA1 1204, or STA21206 may wait for a polling frame transmitted from the AP 1202 (notillustrated) before transmitting the packet.

[0144] The method may continue with the AP 1202 calculating MU-MIMOweight W.sub.i.sup.t1 and W.sub.i.sup.t2 based on andV.sub.i.sup.r2. The AP 1202 may then implement a linear ornon-linear MU-MIMO precoding algorithm for this weight update1209.

[0145] The method may continue with the AP 1202 transmitting 1230again with the best beamforming weights and each STA 1204, 1206receiving with multiple receive antennas.

[0146] The method may continue with one of the followingalternatives. In a first alternative, as illustrated in FIG. 12,the AP 1202 may transmitW.sub.i.sup.t1s.sub.1+W.sub.i.sup.t1s.sub.2 for Max(M1,M2) times1230. s.sub.1 and s.sub.2 may be orthogonal sequences, and may beknown at AP 1202, STA1 1204 and STA2 1206. The AP 1202 may signalSTA1 1204 and STA2 1206 about the assignment of s.sub.1 ands.sub.2. STA1 1204 may be configured to switch receive antennas totrain the best receive beamforming weight W.sub.i.sup.r1.Similarly, STA2 1206 may train the best receive beamforming weightW.sub.i.sup.r2. STA1 1204 may use the orthogonal sequences toestimate the current signal (via cross correlation with s.sub.1)and the current interference (via cross correlation with s.sub.2).Thus, STA1 1204 may train its receive beamforming vectors bynulling the interference. Similarly, STA2 1206 may use theorthogonal sequences to estimate the current signal (via crosscorrelation with s.sub.2) and the current interference (via crosscorrelation with s.sub.1) Then, STA2 1206 may train its receivebeamforming vectors by nulling the interference.

[0147] In a second alternative, the AP 1202 may transmitW.sub.i.sup.t1s for M1 repetitions. STA1 1204 may switch betweenits M1 antennas to receive W.sub.i.sup.t1s, and at the same time,STA2 may monitor the transmission of W.sub.i.sup.t1s. AP 1202 maytransmit W.sub.i.sup.t2s for M2 repetitions and STA2 may switchbetween its M2 antennas to receive W.sub.i.sup.t2s while STA1 1204monitors the transmission of W.sub.i.sup.t2s.

[0148] The example beamforming training method for BDMA may not besuccessful. If the correlation between two STAs 1204, 1206 is high,the two STAs 1204, 1206 may not be distinguished by beams. This maylead to an unsuccessful beamforming training for BDMA. In thisexample, the AP 1202, and or STAs 1204, 1206, may be configured toprovide information if BDMA may or may not be supported with thecurrent configuration. The STAs 1204, 1206 may include the reportof beamforming gain when they report V.sub.i.sup.r1 andV.sub.i.sup.r2 to the AP 1002. The beamforming gain may be definedas .parallel.y.sub.11s'.parallel..sup.2 and.parallel.y.sub.12s'.parallel..sup.2. Alternatively, or inaddition, the STAs 1204, 1206 may report signal to interferenceratio (SIR) during or at the end of the method. The STAs 1004, 1006may determine or calculate the desired signal strength andinterference signal strength with both example alternatives.

[0149] The BDMA training method may include performing BDMAtraining on each STA 1204, 1206 sequentially. This example methodmay be extended for more than two STAs.

[0150] Examples of the BDMA training method may be standardized.For example, the BDMA training method may be used with IEEE802.11ad. In these examples, a service period (SP) may be a timeperiod scheduled for service from one device to another device. Thetransmission during an SP duration may be scheduled by an AP. TheBDMA training method may be scheduled by the AP if it is allocatedin an SP duration. Examples of the BDMA training method may be usedby modifying the BRP procedures.

[0151] FIG. 13 is a diagram of an example modified BRP procedure toimplement a multi-stage iterative beamforming training method forBDMA 1300. In this example, AP 1302 may transmit a BRP frame 1320that indicates a Transmit BDMA BRP Request. A Transmit BDMA BRPRequest subfield may be defined in BRP Request field, and mayindicate that the BRP frame is for transmit BDMA BRP training.Alternatively, a Transmit BRP Request may be used with TX-TRN-REQ=1to indicate that the BRP frame is for transmit BRP training. Aframe that is utilized for single user beamforming training or BDMAtraining may be indicated implicitly or explicitly in the MAC frameor PHY header.

[0152] STA1 1304 may reply with a Transmit BRP Feedback frame 1322by setting TX-train-response=1. STA1 1304 may also request areceive beamforming training by indicating Receive BRP Request inthe same Transmit BRP Feedback frame 1322 by setting L_RX>0.

[0153] AP 1302 may transmit a Polling frame 1324 to STA2 1304 torequest BRP feedback. This step may be skipped if the frame lengthof BRP feedback frame is fixed and known by all the devices.

[0154] STA2 1306 may reply with a Transmit BRP Feedback frame 1326by setting TX-train-response=1. STA2 1306 may also request areceive beamforming training by indicating Receive BRP Request inthe same Transmit BRP Feedback frame 1326 by setting L_RX>0.

[0155] AP 1302 may transmit a BDMA BRP frame 1328 indicating a BRPtrain response by setting RX-Train-response to 1. A BRP frame withRX-Train-response equal to 1 may include a receive trainingsubfield TRN-R 1330 appended to it. A BDMA BRP frame 1328 mayindicate multiple receivers explicitly or implicitly in PHY headeror MAC body.

[0156] The example in FIG. 13 shows that one AP 1302 may transmitto two STAs 1304, 1306. However, the BDMA transmission may be fromone device to two devices irrespective of whether they are APs orSTAs. Moreover, the AP 1302 may transmit to two or more STAs.

[0157] Examples of the method in FIG. 13 may include BDMAprotection mechanisms. The feedback frame 1322 may include not onlythe best beam, but also the achievable SINR. If after a certainnumber of iterations, the achievable SINR is less than the targetSNR, the BDMA method may be aborted. The number of iterations maybe predetermined, determined statically, dynamically determinedbased on previous methods running, or in another way.

[0158] BDMA grouping may be indicated in some examples. Thefollowing examples may enable an indication of BDMA grouping by oneor more communication devices. Using an SP, the BDMA groupinginformation may be indicated in an allocation field in an ExtendedSchedule Element. The Extended Schedule Element may be transmittedin a Beacon frame. The tuple, Source AID, Destination AID andAllocation ID may uniquely identify the allocation. The Source AIDfield may be set to the AID of the STA an may initiate channelaccess during the SP. The Destination AID field may indicate theAID of a STA that may be expected to communicate with the sourceSTA during the allocation. The Allocation ID may identify anairtime allocation from a Source AID to a Destination AID. WithBDMA transmission, more than one receiver may be indicated. Onemethod may be to group BDMA transmitters and receivers, and assigneach group a unique BDMA ID. Each STA corresponding to a BDMA IDmay be assigned a User Position Array that may be used todistinguish the role of the STA. Therefore, the Destination AID maybe replaced by the BDMA ID for BDMA transmissions. Alternatively,more than one Destination AID may be included in the allocationfield. In this way, the order of the Destination AIDs may imply therole of one or more STAs in the BDMA transmission.

[0159] Since the BDMA transmission may be within the SP time slot,the communication device may not need to signal the BDMAtransmission in the PHY header or MAC header. The MCS levels andLength field for each BDMA receiver may be signaled in PHYHeader.

[0160] FIG. 14 is a diagram of an example PHY layer frame format1400 that may be used in a BDMA transmission. N may represent thenumber of BDMA communication devices that may be signaled in theallocation field in Extended Schedule Element. The example PHYlayer frame format 1400 may include an STF field 1410, one or moreCE fields 1420, a header 1430, and a data field 1440. The one ormore CE fields 1420 may be transmitted with a weight and a Pmatrix. The weight for the one or more CE fields may range from W1to WN, for example, the first field may be transmitted with aweight W1, and the last CE field may be transmitted with a weightWN. The header 1430 and data field 1440 may be transmitted usingBDMA and with all of the weights from W1 to WN.

[0161] Example embodiments may include performing BDMA in acontention based access period. For example, BDMA transmissionprotocols may be used by the communication device. Performing aBDMA transmission in a contention based access period may utilizethe NDP announcement (NDPA) and NDP sequences for beamformingtraining. BDMA transmission may be performed after the NDP sequenceexchanges. Alternatively, or in addition, the BDMA transmission maybe delayed until the BDMA initiator, which may be a STA or an AP,acquires the media again.

[0162] In one example BDMA transmission procedure, one or more ofthe communication devices may be configured to use an NDPA period.In this example, an AP may transmit a message that indicates whichSTAs should participate in BDMA training. The NDPA frame maycontain a STA info field to indicate the individual STAinformation. The NDPA frame may reserve a TXOP until the end ofBDMA beamforming training by setting the duration periodaccordingly. Alternatively, the NDPA frame may reserve a TXOP untilthe end of the BDMA transmission.

[0163] In another example, an NDP period may be configured to allowtraining of transmit antennas at the AP. In this example, the STAsmay perform measurements. In another example, a feedback period maybe configured to allow STAs to take turns to feedback the best beamvectors as well as the achievable SINRs. Moreover, STAs may alsofeedback the measured channels or the calculated transmitbeamforming weight vectors.

[0164] In another example, a receiver training period may beconfigured to allow an AP to set its beamforming vectors. In thisexample, STAs may train their receive antennas.

[0165] In another example, the NDP period, feedback period, andreceiver training period may be repeated for a number ofiterations. A number of stopping criteria may be applied in thisexample. For example, if the achievable SINR meets expectation, theiteration may stop early. In these examples, BDMA transmissions maybegin a certain inter-frame spacing after training is performed.ACK1 and ACK 2 may each be followed by a SIFs duration after BDMAtransmission is performed.

[0166] In another example, one or more communication devices may beconfigured to indicate BDMA grouping. The following is an exampleof indicating grouping. BDMA grouping with contention based accessperiod (CBAP) may be performed by using a BDMA ID. A BDMA IDmanagement frame may be transmitted from an AP to a STA to indicatewhether the STA belongs to one of the BDMA groups and the userposition of the STA. The BDMA ID management frame may contain aMembership Status Array field and a User Position Array field. TheBDMA ID may be included in BDMA related frames, such as BDMAtraining frames, BDMA transmission frames, or other similarframes.

[0167] One or more communication devices may be configured toperform a BDMA transmission method for CBAP that may be similar tothat defined for SP. The BDMA transmission may be performed afterthe BDMA initiator, for example the AP, acquires a TXOP in theCBAP. The PHY layer frame format may be the same as illustratedFIG. 14. The transmission of BDMA in CBAP may not be scheduled bythe AP. Accordingly, the BDMA ID may be included in the PHY header.The number of users or communication devices, N, may be indicatedin the sequence exchange to acquire the TXOP before a BDMAtransmission. Alternatively, it may be implicitly indicated using ashort training field (STF) and/or a channel estimation (CE)field.

[0168] In some examples, the communication devices may beconfigured to perform single user spatial multiplexing. In order toperform spatial multiplexing, both transmitter and receiver mayhave multiple RF chains. FIG. 9 and FIG. 10 are example transceiverconfigurations to perform single user spatial multiplexing. Acommunication device with the transceiver configuration shown inFIG. 9 may be referred to as a Type I device, i.e., where multipleRF chains share the same set of antenna elements. A communicationdevice with the transceiver configuration shown in FIG. 10 may bereferred to as a Type II communication device, i.e., where theantenna elements may be split into sub-groups, and each sub-groupmay correspond to one RF chain.

[0169] The communication devices may be configured to performbeamforming methods for spatial multiplexing. In this example,several beamforming methods may be used to perform spatialmultiplexing transmission between a pair of communication devices.Two types of example beamforming methods may be used. The firstexample method may be referred to as Eigen-Beamforming basedspatial multiplexing. In this example, the initiator/responder mayestimate the channel over the air and calculate beamforming weightsaccordingly. The second example method may be referred to as beamsweep based spatial multiplexing. With this method, both initiatorand responder may transmit and receive using pre-defined beamsectors. The beamforming beams may then be selected from these beamsectors.

[0170] The communication devices may be configured to perform anEigen-Beamforming based spatial multiplexing method, where thecommunication devices may be configured as Type I devices withcalibration. Type I devices may have multiple RF chains sharing thesame set of antennas as discussed in conjunction with FIG. 9. Ifthe communication devices are configured to calibrate the multipletransmit RF chains, the communication devices may determine thatthe multiple RF chains are identical. Examples of non-calibrated oridentical RF chains are discussed below.

[0171] FIG. 15 is a diagram of an example beamforming trainingprocedure 1500 using Eigen-beamforming based spatial multiplexingwhere the Type I communication devices may be configured tocalibrate the multiple transmit RF chains. An iterative examplewith two RF chains at both initiator and responder is shown in FIG.15, however, the method may be extended to any number of RF chains.In this example, the transmitter (STA1) 1502 may have Nt antennaelements, and the receiver side (STA2) 1504 may have Nr antennaelements.

[0172] The beamforming training method may be performediteratively. In each iteration, the transmit beamforming trainingmay be performed and then the receive beamforming training may beperformed. An example of a detailed method for Type I devices withcalibration is described below.

[0173] For iteration i, STA1 1502 may act as an initiator, and maytransmit Nt training sequences 1506 sweeping all the transmitantenna elements. The transmission may be performed using the firsttransmit RF chain (TX1) or the second transmit RF chain (TX2) ofSTA1 1502. In some examples, the two RF chains may be identical ordiffer by a scalar, or, in other examples, the two TX chains may becalibrated. Alternatively, STA1 1502 may also use an orthogonalprecoding matrix to transmit the Nt training sequences.

[0174] STA2 (responder) 1504 may have two receive beamformingweights trained from previous iterations, and may be represented byw.sub.t-1.sup.r1 and W.sub.i-1.sup.r2. If this is the firstiteration of the method, STA2 1504 may randomly select twobeamforming weights, or use Omni weights, or may determine the twobeamforming weights in an alternate method. The first RF chain(RX1) may receive a signal that is the weighted combination ofsignals received from all the receive antenna elements. This weightmay be the first receive beamforming weight W.sub.i-1.sup.r1.Similarly, the second RF chain (RX2) may receive a signal that isthe weighted combination of signals received from all the receiveantenna elements. This weight may be the second receive beamformingweight W.sub.i-1.sup.r2. In FIG. 15, the reception of the signalsis shown in dashed lines as an example illustration of the receiveroperation while receiving a packet. For example, the dashed boxW.sub.i-1.sup.r1 may indicate that the STA1 1504 should use receivebeamforming vector W.sub.i-1.sup.r1 to perform receive beamforming.After conversion to the digital domain, STA2 1504 may estimate theeffective channel by comparing the received sequence with the knowntransmitted sequence. For time slot k, STA2 1504 may estimate twochannels using two RF chains

[ G k .times. 1 G k .times. 2 ] . ##EQU00007##

With Nt time slots, STA2 1504 may receive

[ G 1 .times. 1 G N .times. t .times. 1 G 1 .times. 2 G N .times. t.times. 2 ] . ##EQU00008##

Applying the inverse of the orthogonal precoding matrix, STA2 1504may obtain the channel from Nt transmit antenna elements to two RFchains as

H = [ H 1 .times. 1 H N .times. t .times. 1 H 1 .times. 2 H N.times. t .times. 2 ] . ##EQU00009##

[0175] STA2 1504 may feedback the channel information orbeamforming weights for multiple data streams to STA1 1502. STA21504 may calculate the transmit beamforming weights for spatialmultiplexing and feedback 1508 the weights to STA1 1502. STA2 1504may feedback the channel H to STA1 1502, and STA1 1502 then maydetermine or calculate the transmit beamforming weights 1510.

[0176] In some examples, the transmit beamforming weight method maybe implementation dependent. For example, STA1 1502 and/or STA21504 may use linear or non-linear precoding algorithms.

[0177] The updated transmit beamforming weights for the ithiteration may be denoted as (W.sub.i.sup.t1,W.sub.i.sup.t2). STA11502 may transmit a training sequence Nr times with beamformingweight W.sub.i.sup.t1 1512. STA2 1504 may sweep through Nr receiveantennas 1514, or apply an orthogonal matrix. STA2 1504 then passesthe received signal through the two RF chains. STA1 1502 maytransmit training sequences again Nr times with beamforming weightW.sub.i.sup.t2 1516. STA2 1504 may repeat a similar procedure withboth RF chains 1518. The sweeping of the Nr receive antennas 1514and 1518 is shown in dashed lines as an example illustration ofreceiver operation. For example, to receive Nr packets from thetransmitter, the receiver may receive a first packet with the firstreceive antenna, a second packet with the second receive antenna,and so on, until it receives a last packet with the last receiveantenna. STA2 1504 may estimate the channel and update the receivebeamforming weight accordingly (not shown). The receive beamformingweight method may be implementation dependent.

[0178] The above method may be repeated until the method convergesor certain criteria have been met that indicate that spatialmultiplexing is not suitable for the pair of devices, for example,a set of failure criteria). There may be several ways to definefailure criteria that indicate that the pair of devices are notsuitable for spatial multiplexing. A first example of failurecriteria may be that STA2 monitors the rank or condition number ofa channel matrix while selecting a beamforming weight, and mayfeedback this information to STA1. A second example of failurecriteria may be that STA2 monitors the rank or condition number ofchannel matrix while sweeping through Nr receive antennas orapplying an orthogonal matrix, and feeds back this information toSTA1. If the rank is less than the number of data streams expectedto be supported, or the condition number is greater than a certainthreshold, both STA1 and STA2 may determine that the maximum numberof data streams supported may not meet the desired number. In thisexample, the pair of devices may determine to complete the trainingprocedure, and perform RF selection at both transmitter andreceiver later. For example, the pair of devices may terminate thetraining with a full set of RF chains, and return to performbeamforming training with a lesser number of RF chains. Forexample, after training, the devices may transmit with a lessernumber of spatial streams.

[0179] A method of Eigen-beamforming for spatial multiplexing maybe performed. The method may be for communication devices that areof Type I or Type II as discussed in conjunction with FIG. 9.

[0180] Type I devices may perform transmit RF chain trainingsequentially if the RF chains are not calibrated. Even though theRF chains share the same set of antenna elements and the physicalchannels over the air may be the same, the effective channels,which may be the combination of channel over the air andtransmit/receive RF chains, may be measured and estimated.

[0181] Type II devices may split the antenna elements intosub-groups, and each sub-group has an RF chain. In these examples,there may be two RF chains and two sub-groups of antennas. Thephysical channel corresponding to RF chain 1 may be transmittedwith antenna sub-group I, which may be different from thatcorresponding to RF chain 2 that may be transmitted with the otherantenna sub-group. Because of this, the training for multipletransmit RF chains may be performed sequentially.

[0182] FIG. 16 is a diagram of an example beamforming trainingmethod 1600 for Eigen-beamforming based spatial multiplexing forType I devices and for Type II devices without calibration. Foriteration i, STA1 1602 (initiator) may transmit Nt repetitions oftraining sequences sweeping all the transmit antenna elements in afirst antenna sub-group using the first transmit RF chain (TX1)1606. Then STA1 1602 may transmit Nt repetitions of trainingsequences sweeping all the transmit antenna elements in the secondantenna sub-group using the first transmit RF chain (TX2) 1608.STA1 1602 may also use an orthogonal precoding matrix to transmitthe Nt repetitions of training sequences. The first antennasub-group may be the same as the second antenna sub-group for TypeI devices, STA1 1602; while for Type II devices, STA1 1602, theymay correspond to different antenna elements.

[0183] STA2 (responder) 1604 may have the two receive beamformingweights trained from the previous iterations. If this is the firstiteration, STA2 1604 may randomly select two beamforming weights,use Omni weights, or select the weights in an alternate manner. Thefirst receive RF chain (RX1) may obtain a signal as the weightedcombination of signals received from all antenna elements. Theweight may be the first receive beamforming weightw.sub.i-1.sup.r1. Similarly, the second receive RF chain (RX2) mayobtain a signal as the weighted combination of signals receivedfrom all the antenna elements. The weight may be the second receivebeamforming weight W.sub.i-1.sup.r2. The sweeping of the Nr receiveantennas 1616 and 1618 is shown in dashed lines as an exampleillustration of receiver operation. For example, to receive Nrpackets from the transmitter, the receiver may receive a firstpacket with the first receive antenna, a second packet with thesecond receive antenna, and so on, until it receives a last packetwith the last receive antenna. After converting them to basebandand digital domain, STA2 1604 may estimate the effective channel bycomparing the received sequence with the known transmittedsequence. For time slot k, STA2 1604 may estimate two channelsusing two RF chains

[ G k .times. 1 G k .times. 2 ] . ##EQU00010##

With 2Nt time slot, STA2 may receive

[ G 1 .times. 1 G 2 .times. N .times. t .times. 1 G 1 .times. 2 G 2.times. N .times. t .times. 2 ] .times. . ##EQU00011##

The first half of the G matrix,

G T .times. X .times. 1 = [ G 1 .times. 1 G N .times. t .times. 1 G1 .times. 2 G N .times. t .times. 2 ] , ##EQU00012##

may correspond to TX1, and the second half of the G matrix,

G T .times. X .times. 2 = [ G ( Nt + 1 ) .times. 1 G 2 .times. N.times. t .times. 1 G ( Nt + 1 ) .times. 2 G 2 .times. N .times. t.times. 2 ] , ##EQU00013##

may correspond to TX2. Applying the inverse of the orthogonalprecoding matrix to the first half and second half of G matrixrespectively, STA2 1604 may obtain the channel from Nt transmitantenna elements with two transmit RF chains to two receive RFchains

H = [ H 1 .times. 1 H 2 .times. N .times. t .times. 1 H 1 .times. 2H 2 .times. N .times. t .times. 2 ] . ##EQU00014##

[0184] STA2 1604 may transmit channel information or beamformingweights for multiple data streams 1610 to STA1 1602. STA2 1604 maycalculate the transmit beamforming weights to perform spatialmultiplexing for STA1 1602, and transmit the weights 1610 to STA11602. STA2 1604 may transmit the channel H 1610 to STA1 1602, andSTA1 1602 then may determine or calculate the transmit beamformingweights for itself 1620.

[0185] For example, the transmit beamforming weight method may beimplementation dependent, and linear or non-linear precodingmethods may be used. The updated transmit beamforming weights forith iteration may be denoted as (W.sub.i.sup.t1,W.sub.i.sup.t2).STA1 1602 may transmit training sequences Nr times with beamformingweight W.sub.i.sup.t1 1612. STA2 1604 may sweep through Nr receiveantennas 1616, or apply an orthogonal matrix. STA2 1604 may passthe received signal to two RF chains. STA1 1602 may transmittraining sequences 1614 again for Nr times with beamforming weightW.sub.i.sup.t2. STA2 1604 may repeat the same procedure with bothRF chains 1618. The sweeping of the Nr receive antennas 1616 and1618 is shown in dashed lines as an example illustration ofreceiver operation. For example, to receive Nr packets from thetransmitter, the receiver may receive a first packet with the firstreceive antenna, a second packet with the second receive antenna,and so on, until it receives a last packet with the last receiveantenna. STA2 1604 may estimate the channel and update the receivebeamforming weight 1622 accordingly.

[0186] For example, the receive beamforming weight method may beimplementation dependent, and may be repeated for severaliterations until the algorithm converges or certain criteria havebeen met that indicate that spatial multiplexing is not suitablefor the pair of devices, STA1 1602 and STA2 1604.

[0187] There may be several ways to define failure criteria toindicate that the pair of devices are not suitable for spatialmultiplexing. For example, the failure criteria may include STA2monitoring the rank or condition number of channel matrix whenselecting beamforming weights, and feeding back this information toSTA1. A second example of a failure criteria may include STA2monitoring the rank or condition number of channel matrix whilesweeping through Nr receive antennas or applying an orthogonalmatrix, and feeding back this information to STA1.

[0188] If the rank is less than the number of data streams expectedto be supported, or the condition number is greater than a certainthreshold, both STA1 and STA2 may determine that the maximum numberof data streams that may be supported does not meet therequirements. In this example, the pair of devices may determine tocomplete the training procedure, and perform RF selection at bothtransmitter and receiver later. Alternatively, the pair of devicesmay terminate the training with full set of RF chains, and returnto performing beamforming training with a fewer number of RFchains. After training, they may transmit with a fewer number ofspatial streams.

[0189] In some examples, methods for beam sweep based spatialmultiplexing for Type I devices with calibration may be performed.For example, the method of beam sweep based spatial multiplexingmay be similar to Eigen-Beamforming based spatial multiplexing.Examples using Eigen-Beamforming based spatial multiplexing mayrequire that the channel estimate and the transmit/receive weightsfor spatial multiplexing may be determined based on the estimatedchannel, which may not necessarily be the same as one of the beamsused for beamforming training. In beam sweep based spatialmultiplexing, there may be no requirement for channel estimation.The device may select one or multiple beams from the set of beamsused for beam sweep training. For example, implementation of beamsweep based beamforming may be easier than Eigen-Beamforming basedbeamforming. The performance of the beam sweep based methods may besub-optimum compared to the Eigen-beamforming based methods.

[0190] FIG. 17 is a diagram of an example beamforming trainingmethod 1700 for beam sweep based spatial multiplexing for Type Idevices with calibration between two TX chains. Referring to FIG.17, for iteration i, STA1 (initiator) 1702 may transmit Nrepetitions of training sequences 1706 sweeping the transmit beamsit intends to train. In these examples, N may not necessarily berelated to the number of transmit antennas. The transmission 1706may be performed using the first transmit RF chain (TX1) or thesecond transmit RF chain (TX2). The two RF chains may be identicalor different by a scalar. In some examples, the two TX chains mayhave been calibrated.

[0191] STA2 (responder) 1704 may have the two receive beams trainedfrom the previous iterations. If this is the first iteration, STA21704 may randomly select two beams, use Omni weights, or selectinitial values in an alternate manner. The first receive RF chain(RX1) of STA2 1704 may obtain a signal as the weighted combinationof signals received from all antenna elements. The weight may bethe first receive beamforming weight W.sub.t-1.sup.r1. Similarly,the second receive RF chain (RX2) of STA2 1704 may obtain a signalas the weighted combination of signals received from all theantenna elements. The weight may be the second receive beamformingweight W.sub.i-1.sup.r2. {W.sub.i-1.sup.r1,W.sub.i-1.sup.r2} may bethe weights corresponding to beam indices{ID.sub.i-1.sup.r1,ID.sub.i-1.sup.r2}. After converting them tobaseband and digital domain, STA2 1704 may measure the effectiveSNR or equivalent parameters. For time slot k, STA2 1704 mayperform SNR measurements using two receive RF chains

[ S .times. N .times. R k .times. 1 S .times. N .times. R k .times.2 ] . ##EQU00015##

With N time slots, STA2 1704 may receive

[ S .times. N .times. R 1 .times. 1 S .times. N .times. R N .times.1 S .times. N .times. R 1 .times. 2 S .times. N .times. R N .times.2 ] . ##EQU00016##

In FIG. 17, the reception of the signals is shown in dashed linesas an example illustration of the receiver operation whilereceiving a packet.

[0192] STA2 1704 may feedback two beam indices to STA1 1702. Thebeam selection method may be implementation dependent. For example,the STA2 1704 may choose the pair of indices(ID.sub.i.sup.t1,ID.sub.i.sup.t2) which may satisfyID.sub.i.sup.t1=arg max.sub.k(SNR.sub.k1.sup.TX-SNR.sub.k2.sup.TX),and ID.sub.i.sup.t2=argmax.sub.k(SNR.sub.k2.sup.TX-SNR.sub.k1.sup.TX).

[0193] The updated transmit beam indices for the ith iteration maybe (ID.sub.i.sup.t1, ID.sub.i.sup.t2) STA1 1702 may transmit atraining sequence for M times with beam ID.sub.i.sup.t2 1720. STA21704 may sweep through M receive beams with both receive RF chains(RX1 and RX2) 1722. STA1 1702 may transmits a training sequenceagain for M times with beamforming weight ID.sub.i.sup.t2 1724.STA2 1704 may repeat the similar procedure with both RF chains1726. The sweeping of the receive antennas 1722 and 1726 is shownin dashed lines as an example illustration of receiver operation.For example, to receive packets from the transmitter, the receivermay receive a first packet with the first receive antenna, a secondpacket with the second receive antenna, and so on, until itreceives a last packet with the last receive antenna. STA2 1704 maymeasure the SNR or equivalent parameters and update the receivebeam index accordingly. The receive beam selection method may beimplementation dependent. For example, M may be the number ofreceive beams STA2 1704 intends to train and it may not necessarilybe related to a number of receive antennas at STA2 1704. The methodmay be repeated until the method converges or certain criteria havebeen met that indicates that spatial multiplexing is not suitablefor the pair of devices.

[0194] Failure criteria may be defined in several different ways.For example, failure criteria may indicate that the pair of devicesis not suitable for spatial multiplexing. In one example, thefailure criteria may be defined as when STA2 1704 may record.DELTA..sub.11.sup.SNR=max.sub.k(SNR.sub.k1.sup.TX-SNRk.sub.k2.sup.TX)and.DELTA..sub.12.sup.SNR=max.sub.k(SNR.sub.k2.sup.TX-SNR.sub.k1.sup.TX)when selecting beams, and feedback this information to STA1 1702.In another example, STA2 1704 may record.DELTA..sub.21.sup.SNR=max.sub.1.ltoreq.k.ltoreq.N(SNR.sub.k1.sup.RX-SNR.-sub.k2.sup.RX) and.DELTA..sub.22.sup.SNR=max.sub.N<k.ltoreq.2N(SNR.sub.k2.sup.RX-SNR.sub-.k1.sup.RX) when sweeping beams, and feedback this information toSTA1 1702.

[0195] If .DELTA..sub.ij.sup.SNR is smaller than a certainthreshold, both STA1 1702 and STA2 1704 may determine that thechannel cannot provide enough spatial diversity to support two datastreams. In this example, STA1 1702 and STA2 1704 may determine tocomplete the training procedure, and perform RF selection at bothtransmitter and receiver. STA1 1702 and STA2 1704 may terminate thetraining with two RF chains, and return to performing beamformingtraining with one RF chain. After training, they may transmit witha fewer number of spatial streams. For example, more than two datastreams may be determined.

[0196] Beam sweep based spatial multiplexing for Type II devicesand Type I devices without calibration may be performed. FIG. 18 isa diagram of an example beamforming training method 1800 for beamsweep based spatial multiplexing Type II devices and Type I deviceswithout calibration.

[0197] For iteration i, STA1 (initiator) 1802 may transmit Nrepetition of training sequences sweeping all the transmit beams itintends to train using the first transmit RF chain (TX1) 1806. ThenSTA1 1802 may repeat the same procedure with the second RF chain(TX2) 1808. N may not necessarily be related to the number oftransmit antennas. For example, the beam pattern used for TX1 maynot be the same as that for TX2.

[0198] STA2 (responder) 1804 may have the two receive beams trainedfrom the previous iterations. If this is the first iteration, STA21804 may randomly select two beams, use Omni weights, or select thetwo beams in a different way. The first receive RF chain (RX1) 1810may obtain a signal as the weighted combination of signals receivedfrom all antenna elements. The weight may be the first receivebeamforming weight W.sub.i-1.sup.r1. The second receive RF chain(RX2) 1812 may obtain a signal as the weighted combination ofsignals received from all the antenna elements. The weight may bethe second receive beamforming weight W.sub.i-1.sup.r2. Forexample, {W.sub.i-1.sup.r1,W.sub.i-1.sup.r2} may be the weightsthat correspond to beam indices{ID.sub.i-1.sup.r1,ID.sub.i-1.sup.r2}. After converting them tobaseband and digital domain, STA2 1804 may measure the effectiveSNR or equivalent parameters. For time slot k, STA2 1804 mayperform SNR measurements using two receive RF chains

[ S .times. N .times. R k .times. 1 S .times. N .times. R k .times.2 ] . ##EQU00017##

With the first N time slot, STA2 1804 may receive

[ S .times. N .times. R 1 .times. 1 S .times. N .times. R N .times.1 S .times. N .times. R 1 .times. 2 S .times. N .times. R N .times.2 ] , ##EQU00018##

which may correspond to TX1 of STA1 1802. With the last N timeslots, STA2 1804 may receive

[ S .times. N .times. R ( N + 1 ) .times. 1 S .times. N .times. R 2.times. N .times. 1 S .times. N .times. R ( N + 1 ) .times. 2 S.times. N .times. R 2 .times. N .times. 2 ] , ##EQU00019##

which may correspond to TX2 of STA1 1802. In FIG. 18, the receptionof the signals is shown in dashed lines as an example illustrationof the receiver operation while receiving a packet.

[0199] STA2 1804 may feedback two beam indices 1814 to STA1 1802.The beam selection method may be implementation dependent. Forexample, the STA2 1804 may select the pair of indices(ID.sub.i.sup.t1,ID.sub.i.sup.t2) which may satisfyID.sub.i.sup.t1=argmax.sub.1.ltoreq.k.ltoreq.N(SNR.sub.k1.sup.TX-SNR.sub.k2.sup.TX)and ID.sub.i.sup.t2=argmax.sub.N<k.ltoreq.2N(SNR.sub.k2.sup.TX-SNR.sub.k1.sup.TX).

[0200] The updated transmit beam indices for ith iteration may be(ID.sub.i.sup.t1,ID.sub.i.sup.t2). STA1 1802 may transmit atraining sequence M times with beam ID.sub.i.sup.t1 1816. STA2 1804may sweep through M receive beams with both receive RF chains (RX1and RX2) 1818. STA1 1802 may transmit a training sequence again Mtimes with beamforming weight ID.sub.i.sup.t2 1820. STA2 1804 mayrepeat the same procedure with both RF chains 1822. The sweeping ofthe receive antennas 1818 and 1822 is shown in dashed lines as anexample illustration of receiver operation. For example, to receivepackets from the transmitter, the receiver may receive a firstpacket with the first receive antenna, a second packet with thesecond receive antenna, and so on, until it receives a last packetwith the last receive antenna. STA2 1804 may measure the SNR orequivalent parameters and update the receive beam indexaccordingly. The receive beam selection method may beimplementation dependent. For example, M may be the number ofreceive beams STA2 1804 intends to train and may not necessarily berelated to the number of receive antennas at STA2 1804.

[0201] The procedure may be repeated for several iterations untilthe method converges or certain criteria have been met thatindicates that spatial multiplexing is not suitable for STA1 1802and STA2 1804.

[0202] There may be several ways to define failure criteria thatindicate that the pair of devices are not suitable for spatialmultiplexing. For example, the failure criteria may be defined aswhen STA2 1804 may record.DELTA..sub.11.sup.SNR=max.sub.1.ltoreq.k.ltoreq.N(SNR.sub.k1.sup.TX-SNR.-sub.k2.sup.TX) and.DELTA..sub.12.sup.SNR=max.sub.N<k.ltoreq.2N(SNR.sub.k2.sup.TX-SNR.sub-.k1.sup.TX) in when selecting beams, and feedback this informationto STA1 1802. In another example failure criteria, STA2 1804 mayrecord.DELTA..sub.21.sup.SNR=max.sub.1.ltoreq.k.ltoreq.N(SNR.sub.k1.sup.RX-SNR.-sub.k2.sup.RX) and.DELTA..sub.22.sup.SNR=max.sub.N<k.ltoreq.2N(SNR.sub.k2.sup.RX-SNR.sub-.k1.sup.RX) in when sweeping beams, and feedback this informationto STA1 1802.

[0203] If .DELTA..sub.ij.sup.SNR is smaller than a threshold, bothSTA1 1802 and STA2 1804 may determine that the channel cannotprovide enough spatial diversity to support two data streams. Inthis example, the pair of devices may determine to complete thetraining procedure, and perform RF selection at both transmitterand receiver. Alternatively, the pair of devices may terminate thetraining with two RF chains, and return to performing beamformingtraining with one RF chain. After training, the pair of devices maytransmit with a fewer number of spatial streams.

[0204] The beam refinement transaction discussed in conjunctionwith FIG. 3 may be used for Eigen-beamforming based spatialmultiplexing methods disclosed above. Modifications may be appliedto support spatial multiplexing. For example, a number of spatialstreams may be defined. The number of data streams may be definedin a DMG beam refinement element. The FBCK-TYPE subfield in the DMGbeam refinement element may be modified.

[0205] FIG. 19 is a diagram of an example modified FBCK-TYPEsubfield 1900. The modified FBCK-TYPE subfield 1900 may be includedin a DMG refinement element. The modified FBCK-TYPE subfield 1900may include a SNR present field 1910, a channel measurement presentfield 1920, a tap delay present field 1930, a number of tapspresent field 1940, a number of measurement field 1950, a number ofspatial streams field 1960, a sector ID order present field 1970,and a number of beams field 1980.

[0206] An initiator may determine the capabilities of the responderprior to initiating beamforming training with the responder byusing an Eigen-beamforming based spatial multiplexing capability.The Beam sweep based spatial multiplexing capability may beindicated in a DMG capabilities element. The DMG capabilitieselement may be present in an association request, associationresponse, re-association request, re-association response, proberequest and probe response frames and may be present in DMG beaconand information request and response frames. One bit ofEigen-beamforming based spatial multiplexing indication and one bitof beam sweep spatial multiplexing capability may be used toindicate that the STA is capable of performing Eigen-beamformingbased spatial multiplexing.

[0207] The type of beamforming training algorithm, such asEigen-beamforming based and beam sweep based, may be indicated inDMG beam refinement element. In addition, transceiver architecturetype, such as Type I and Type II may be indicated in a DMGcapabilities element.

[0208] A precoding matrix utilized by the initiator in the firstpart of each iteration may be predefined and agreed on by bothinitiator and responder if Eigen-beamforming based spatialmultiplexing is implemented. In this example, the initiator andresponder may negotiate which matrix to utilize before thebeamforming training. For example, the precoding matrix index maybe defined and transmitted in a BRP request field. In addition, aset of unitary precoding matrices may be predetermined.

[0209] A number of antennas at both initiator and responder may besignaled if Eigen-beamforming based spatial multiplexing isimplemented. The number of antennas may be signaled in the PHYheader, MAC header or a BRP Request field.

[0210] A spatial multiplexing frame format may be implemented. Forexample, when a packet is transmitted using spatial multiplexing,an indication may be sent to inform the packet recipients thatmultiple streams were transmitted. The MCS may be redefined for amodulation/coding scheme and the number of spatial streams. In802.11ad, for example, MCS 0 may be the Control PHY; MCS 1-12 maybe utilized for single carrier (SC) PHY; MCS 13-24 may be for OFDMPHY; and MCS 25-31 may be for low power SC PHY.

[0211] In examples with two data stream transmissions, thefollowing may be defined for use by the communication devices. Forexample, MCS 32-43 may be for SC PHY, MCS 44-55 may be for OFDMPHY, and 56-62 may be for low power SC PHY. In some examples, theMCS mapping may not be the same as defined above.

[0212] Alternatively, the number of spatial streams may beindicated in a PHY header. In order to support multiple datastreams, the PHY layer frame format may need to be modified.

[0213] FIGS. 20A, 20B, and 20C are diagrams of example PHY layerframe formats. With SC PHY, the data field may be composed ofsymbol blocks, while with OFDM PHY, the data field may be composedof OFDM symbols. The frame may be appended with TRN-T/R subfields,and may be utilized for beam refinement protocol.

[0214] Referring to FIG. 20A, short training field (STF) 2010,channel estimation field (CE) 2020 and PHY headers 2030 may betransmitted with a weight, W1 2040. The number of data streamssupported, N, may be indicated in the PHY header. If more than onedata stream will be transmitted, additional CE field(s) may beincluded. With N data streams, an extra N-1 CE field 2050 may betransmitted and weights W2 2060, . . . , WN 2070 may be applied toeach CE field. An orthogonal mapping matrix, such as the P matrixdefined in 802.11n/ac, may be applied. If a cyclic shift delay(CSD) scheme is applied to spatial multiplexing, the same CSDparameters may be applied to the CE fields. The data field 2080that follows may be transmitted using the spatial multiplexingscheme, and all of the weights (W1, . . . , WN) 2090 may beapplied.

[0215] FIG. 20B is a diagram of another example preamble format forspatial multiplexing transmissions. This format is similar to FIG.20A except that an AGC field 2015 may be inserted after additionalCE fields and before the data field 2080. The AGC field 2015 mayuse the same sequence as an LTF field, or it may be redesigned. Thepurpose of this AGC field 2015 may be for automatic gain control.The transmission of the AGC field 2015 may be in the same format asthe data field 2080, i.e., weights (W1, . . . , WN) 2090 may beapplied. The same CSD parameters may be applied to AGC field 2015if CSD is utilized for data transmissions.

[0216] FIG. 20C is a diagram of another example preamble format forspatial multiplexing transmissions in which the number of datastreams may be signaled implicitly. STF 2010 may be transmittedusing all the weights (W1, . . . , WN) 2025. The first CE field2020 following STF 2010 may be transmitted using the first weightW1 2040. The number of data streams may be implicitly indicated byusing STF 2010 and the first CE field 2020. For example, several CEsequences may be defined, and each sequence may correspond to acertain number of data streams. Additional N-1 CE fields 2050 mayfollow the first CE field 2020 and transmitted with weights W2 2060to WN 2070. The header 2035 may be transmitted with one of theweights or a combination of the weights similar to STF 2010. Thespatial multiplexing transmission may be transmitted following theadditional CE fields.

[0217] Beamforming training overhead and latency may be reduced.For example, sector sweep (SSW) frames and related training methodsmay be modified.

[0218] In the SLS procedures, SSW frames may be utilized fortransmit and receive beamforming training. For example, the SSWframes may be transmitted in N time slots. For transmit beamformingtraining, SSW frames may be transmitted and multiple antennasectors may be swept. The receiver may receive the SSW frames withthe same antenna sector and feedback the best transmit sector ID tothe transmitter. For example, for receive beamforming training, thesame SSW frames may be repeated N times, and the receiver may sweepover multiple antenna sectors to receive. After the receivebeamforming training, the receiver may select the best receivesector.

[0219] Each SSW frame may comprise a full PLCP header that mayinclude a preamble, one or more header blocks, and a MAC frame.Since the SSW frames may be utilized for beamforming training, theymay be transmitted using the lowest data rate, for example, controlPHY or MCSO in 802.11ad. SSW frames may not contain data traffic,therefore SSW frame sequences may be beamforming trainingoverhead.

[0220] FIG. 21 is a diagram of an example modified SSW trainingframes and sequence 2100. In this example, modified SSW trainingsequences may be utilized. A SSW announcement (SSWA) frame 2110 maybe transmitted at the beginning of the SSW training sequences. TheSSWA frame 2110 may contain all the information used to transmit bySSW frames. One or more N null SSW (NSSW) frames 2120 may followthe SSWA frame 2110 with a certain inter-frame spacing. NSSW frames2120 may contain only preamble and PHY headers, and no MACframe.

[0221] FIG. 22 is a diagram of an example SSWA frame format 2200.The SSWA frame format 2200 may include a frame control field 2205,a duration field 2210, an RA field 2215, a TA field 2220, an SSWfield 2225, an SSW feedback (FB) field 2230, and an FCS field 2235.The SSW field 2225 may include a direction subfield 2240, a DMGantenna ID 1 subfield 2245, a sector ID 1 subfield 2250, a sectorID N subfield 2255, a DMG antenna ID 2 subfield 2260, a sector ID 1subfield 2265, a sector ID N2 subfield 2270, and an RXSS lengthsubfield 2275. In this example, sector ID 1 subfield 2250 may befor DMG antenna ID 1, and sector ID 1 subfield 2265 may be for DMGantenna ID 2.

[0222] The direction subfield 2240 and the RXSS length subfield2275 may be the same as in IEEE 802.11ad. The direction subfield2240 may be set to 0 to indicate that the frame is transmitted bythe beamforming initiator and set to 1 to indicate that the frameis transmitted by the beamforming responder. The RXSS Lengthsubfield 2275 may be valid only when transmitted in a CBAP and maybe reserved otherwise. The RXSS Length subfield 2275 may specifythe length of a receive sector sweep as required by thetransmitting STA, and may be defined in units of an SSW frame. Thevalue of this field is in the range 0-62, with odd values beingreserved.

[0223] DMG Antenna IDs and Sector IDs may be utilized to indicatethe antenna pattern for the following NSSW frames. For example, thefirst NSSW frame may utilize DMG Antenna ID 1 and Sector ID 1 totransmit, and the second NSSW frame may utilize DMG Antenna ID 1and Sector ID 2 to transmit, and so on. With DMG antenna ID k,there may be Nk sectors swept for this round of beamformingtraining. The total number of NSSW frames following this SSWA framemay be

k = 1 K .times. N k . ##EQU00020##

K may be the number of DMG antennas trained with these SSWA-NSSWsequences.

[0224] The SSWA may be transmitted as follows. For example, theSSWA frame may carry all the MAC information necessary forbeamforming training. It may be important that the receiver decodesthe SSWA frame correctly. The SSWA frame may be transmitted usingone of the following methods. For example, if the beamformingtraining is between two non-AP/PCP devices, the SSWA frame may betransmitted from AP to the two devices. In another example, if boththe beamforming initiator and responder are multi-band capable,they may operate on multiple frequency bands simultaneously, andthe SSWA frame may be transmitted on another frequency band. TheSSWA frame may be transmitted with low data rate, spreading codes,or repetition schemes.

[0225] Some examples may use sub-optimum SLS training methods. Inthese examples, SLS training methods may be terminated early.

[0226] FIG. 23 is a diagram of an example early termination of theSLS training procedure. In this example, the initiator 2302 mayhave 4 beam sectors to train, and the countdown (CDOWN) number mayequal 3 in the first training frame. In this example, sector 3 isshaded to illustrate that the initiator may use sector 3 fortransmission in a first period, and sector 4 is shaded toillustrate that the initiator may use sector 4 for transmission inthe second period. The initiator 2302 may continue transmittingtraining frames which are separated by inter-frame space duration2T, for example. The responder 2304 may monitor the receivedtraining frames 2306. The dashed "omni" circles in FIG. 23 areshown to illustrate that the receiver/responder may be in anomni-receiving mode. An omni-receiving mode may be enabled by anomni-directional receiving antenna. The first two "omni" circlesare shown in dashed lines to illustrate that they are examplereceiver operations. The last "omni" circle is shown in indicatingsolid line to illustrate that this is an example transmitteroperation, i.e., the feedback packet may be transmitted in anomni-transmitting mode, which may be enabled by an omni-directionaltransmitting antenna. Once the received SNR (or other parameters)is greater than a certain threshold, the responder 2304 maydetermine to terminate the training procedure by transmitting afeedback frame 2308. The feedback frame 2308 may be transmittedafter a T duration from the end of a training frame transmitted bythe initiator 2302. Thus the initiator 2302 may detect thetransmission of this feedback and stop transmitting more trainingframes. This example may be used for both transmit and receivebeamforming training.

[0227] A group based SLS training method may be performed. In thisexample, a STA may divide its sectors to groups. The partition ofsectors may be implementation dependent. For example, the partitionmay be based on the direction of the sectors. The beamforminginitiator may select one group to perform SLS training and wait forthe feedback from the responder. Once the feedback from respondermeets the expectation of the initiator, the beamforming initiatormay determine to stop the beamforming training. Otherwise, theinitiator may select another group to perform SLS training untilone beam is selected or all the beams are swept.

[0228] Multi-beam, multi-DMG antenna sector level sweep feedbackmay be performed. In examples of the sector level sweep method, thereceiver STA may report back the best beam only, for example, onesector of one DMG antenna. For example, a list of the best beams,such as multiple beams of multiple antennas may be reported. Thismay enable the communication link to track the relative performanceof the beams over time and, if necessary, switch to a better beamwithout the need for retraining.

[0229] FIG. 24 is a diagram of an example multi-beam multi-DMGantenna SLS feedback method 2400. The following method 2400 may usea sector level sweep method that may report a list of best beams.In this example, the DMG transmitting STA1 2402 and receiving STA22404 may indicate their capability to support multi-sector,multi-DMG antenna Sector Sweep (SSW) feedback. This capability maybe indicated by a bit in the DMG STA Capability Information field.STAs that do not have the capability may fall back to legacytransmission.

[0230] The STA initiating the sector sweep, for example STA1 2402,may transmit information to the responder STA, for example STA22404, indicating the number of beams to be fed back. The responderSTA 2404 may also transmit information to initiator STA 2402 on thenumber of beams to be fed back. The metric to decide on the bestbeams may be implementation dependent. One signaling method may usea Transmit Sector Sweep frame for both the initiator and theresponder may contain the number of beams to be fed back. A secondsignaling method may use a DMG beacon that may contain a field thatindicates the number of beams to be fed back for all SSW feedback2420. In a third signaling method, before a Sector Level Sweepprocedure, the initiator and responder may exchange SLS setupframes indicating the number of beams to feed back. Quasi-omni mayrefer to a near omni-directional transmission or reception. Forexample, quasi-omni transmissions may be enabled by repeatedlytransmitting the same information using multiple directionaltransmissions, as if it were transmitted using an omni-directionaltransmit antenna. Similarly, quasi-omni receptions may be enabledby repeatedly receiving the same packet using multiple directionalreceptions, as if it were received using an omni-directionalreceiving antenna.

[0231] Both STAs may implement the legacy initiator and respondersector level sweep procedures. The transmitter may feedback thebest N beams. This may be by one of the following example methods.In a first example method, multiple SSW Feedback fields may beaggregated within an SSW feedback frame 2420. In a second example,a single SSW Feedback field may be modified to enable feedback ofmultiple beams and DMG antennas and corresponding SNR Reports. In athird example, the best beam/antenna may be fed back during the SLSprocedure and subsequent feedback of the additional N-1 beams withother transmissions, for example an ACK. In these examples, thenumber of antenna may be larger than the number radio frequency(RF) chains. In some examples, the number of antenna may be muchlarger than the number of RF chains.

EMBODIMENTS

[0232] 1. A first communication device comprising: [0233] aplurality of antennas; [0234] a processor configured to partitionthe plurality of antennas; [0235] a transmitter configured totransmit a plurality of frames to a second communication device;[0236] and a receiver.

[0237] 2. The first communication device of embodiment 1, whereinthe processor is configured to partition the plurality of antennasinto at least a first group of antennas and a second group ofantennas.

[0238] 3. The first communication device of embodiment 2, whereinthe first group of antennas is associated with a first beam to afirst station (STA).

[0239] 4. The first communication device of embodiment 2 or 3,wherein the second group of antennas is associated with a secondbeam to a second station (STA).

[0240] 5. The first communication device of any one of embodiments2-4, wherein the first group of antennas is associated with a firstbeam to a first plurality of stations (STAs).

[0241] 6. The first communication device of any one of embodiments2-5, wherein the second group of antennas is associated with asecond beam to a second plurality of stations (STAs).

[0242] 7. The first communication device of any precedingembodiment, wherein the plurality of frames transmitted arebeamforming training frames.

[0243] 8. The first communication device of any precedingembodiment, wherein the plurality of frames are transmitted usingthe first group of antennas.

[0244] 9. The first communication device of any precedingembodiment, wherein the plurality of frames are transmitted usingthe second group of antennas.

[0245] 10. The first communication device of any precedingembodiment, wherein the receiver is configured to receive a firstbeamforming weight vector from the second communication device.

[0246] 11. The first communication device of embodiment 10, whereinthe first beamforming weight vector is for sending signals on thefirst group of antennas.

[0247] 12. The first communication device of any precedingembodiment, wherein the receiver is configured to receive a secondbeamforming weight vector from the second communication device.

[0248] 13. The first communication device of embodiment 12, whereinthe second beamforming weight vector is for sending signals on thesecond group of antennas.

[0249] 14. The first communication device of any one of embodiments10-13, wherein the first beamforming weight vector is a strongestbeam between the first communication device and the secondcommunication device.

[0250] 15. The first communication device of any one of embodiments12-14, wherein the second beamforming weight vector is for a secondstrongest beam between the first communication device and thesecond communication device.

[0251] 16. The first communication device of any precedingembodiment, wherein the first communication device is a wirelesstransmit/receive unit (WTRU).

[0252] 17. The first communication device of any one of embodiments1-15, wherein the first communication device is a station(STA).

[0253] 18. The first communication device of any one of embodiments1-15, wherein the first communication device is an access point(AP).

[0254] 19. The first communication device of any one of embodiments1-15, wherein the first communication device is a base station.

[0255] 20. The first communication device of any precedingembodiment, wherein the second communication device is a wirelesstransmit/receive unit (WTRU).

[0256] 21. The first communication device of any one of embodiments1-19, wherein the second communication device is a station(STA).

[0257] 22. The first communication device of any one of embodiments1-19, wherein the second communication device is an access point(AP).

[0258] 23. The first communication device of any one of embodiments1-19, wherein the second communication device is a basestation.

[0259] 24. The first communication device of any one of embodiments7-23, wherein the beamforming training frames are orthogonalbeamforming vectors.

[0260] 25. The first communication device of any precedingembodiment, wherein the transmitter is further configured totransmit a second set of beamforming training frames.

[0261] 26. The first communication device of embodiment 25, whereinthe second set of beamforming training frames is transmitted usingthe received first beamforming weight vector.

[0262] 27. The first communication device of embodiment 25 or 26,wherein the second set of beamforming training frames istransmitted using the received second beamforming weightvector.

[0263] 28. The first communication device of any precedingembodiment, wherein the receiver is further configured to receive amodified first beamforming weight vector.

[0264] 29. The first communication device of embodiment 28, whereinthe modified first weight vector is for sending signals on thefirst group of antenna.

[0265] 30. The first communication device of any precedingembodiment, wherein the receiver is further configured to receive amodified second beamforming weight vector.

[0266] 31. The first communication device of embodiment 30, whereinthe modified second weight vector is for sending signals on thesecond group of antenna.

[0267] 32. The first communication device of any precedingembodiment, wherein the first communication device comprises one ormore radio frequency (RF) chains.

[0268] 33. The first communication device of any precedingembodiment, wherein a number of the antenna is larger than a numberof one or more radio frequency (RF) chains.

[0269] 34. A first communication device comprising: [0270] aplurality of antennas; [0271] a receiver configured to receive aset of beamforming training frames; [0272] a processor; and [0273]a transmitter.

[0274] 35. The first communication device of embodiment 34, whereinthe processor is configured to determine a first transmitbeamforming weight vector.

[0275] 36. The first communication device of embodiment 35, whereinthe first transmit beamforming weight vector corresponds to a firstantenna group.

[0276] 37. The first communication device of embodiment 36, whereinthe first antenna group is for a second communication device.

[0277] 38. The first communication device of any one of embodiments34-37, wherein the processor is configured to determine a secondtransmit beamforming weight vector.

[0278] 39. The first communication device of embodiment 38, whereinthe second transmit beamforming weight vector corresponds to asecond antenna group.

[0279] 40. The first communication device of embodiment 39, whereinthe second antenna group is for a second communication device.

[0280] 41. The first communication device of any one of embodiments34-40, wherein the transmitter is configured to transmit data usingthe first transmit beamforming weight vector to the secondcommunication device to the second communication device.

[0281] 42. The first communication device of any one of embodiments34-41, wherein the transmitter is configured to transmit data usingthe second transmit data using the second transmit beamformingweight vector to the second communication device.

[0282] 43. The first communication device of any one of embodiments34-42, wherein the first communication device is a wirelesstransmit/receive unit (WTRU).

[0283] 44. The first communication device of any one of embodiments34-42, wherein the first communication device is a station(STA).

[0284] 45. The first communication device of any one of embodiments34-42, wherein the first communication device is an access point(AP).

[0285] 46. The first communication device of any one of embodiments34-42, wherein the first communication device is a basestation.

[0286] 47. The first communication device of any one of embodiments34-46, wherein the second communication device is a wirelesstransmit/receive unit (WTRU).

[0287] 48. The first communication device of any one of embodiments34-46, wherein the second communication device is a station(STA).

[0288] 49. The first communication device of any one of embodiments34-46, wherein the second communication device is an access point(AP).

[0289] 50. The first communication device of any one of embodiments34-46, wherein the second communication device is a basestation.

[0290] 51. The first communication device of any one of embodiments34-50, wherein the received beamforming training frames areorthogonal beamforming vectors.

[0291] 52. The first communication device of any one of embodiments34-51, wherein the transmitted beamforming weight vectors areorthogonal beamforming vectors.

[0292] 53. A method for beamforming training for beam divisionmultiple access (BDMA), the method comprising: [0293] receiving afirst transmit beamforming weight from a first station (STA), andreceiving a second transmit beamforming weight from a second (STA);[0294] transmitting Nt sequences modulated using Nt beamformingvectors, wherein the Nt sequences are modulated based on the firsttransmit beamforming weight and the second transmit beamformingweight.

[0295] 54. The method of embodiment 53, wherein the Nt beamformingvectors are orthogonal.

[0296] 55. A method for beamforming training for beam divisionmultiple access (BDMA), the method comprising: [0297] receiving aplurality of Nt sequences using a first previous beamformingvector; [0298] determining a first transmit beamforming weight froman access point (AP) based on the first previous beamforming vectorand the received plurality of Nt sequences; and [0299] transmittingthe determined first transmit beamforming weight to the AP.

[0300] 56. A method for beamforming training for spatialmultiplexing, the method comprising: [0301] an initiator stationhaving two radio frequency (RF) chains and having transmit antennaelements, the initiator station transmitting Nt known trainingsequences sweeping the transmit antenna elements; [0302] aresponder station having a first and a second RF chain and havingreceive antenna elements, and having a first and a second receivebeamforming weight, the responder station receiving the first RFchain and receiving the second RF chain; and [0303] the responderstation estimating at least two channels by comparing the receivedfirst RF chain and the second RF chain to the Nt known trainingsequences.

[0304] 57. The method of embodiment 56, wherein the first RF chainis the combination of signals received from the receive antennaelements weighted by the first receive beamforming weights

[0305] 58. The method of embodiment 57, wherein the second RF chainis the combination of signals received from the receive antennaelements weighted by the second receive beamforming weights.

[0306] 59. The method of any one of embodiments 56-58, wherein thetransmitting Nt training sequences further comprises transmittingNt training sequences over a first RF chain of the initiator.

[0307] 60. The method of any one of embodiments 56-58, wherein thetransmitting Nt training sequences further comprises transmittingNt training sequences over a second RF chain of the initiator.

[0308] 61. The method of any one of embodiments 56-58, wherein thetransmitting Nt training sequences further comprises transmittingNt training sequences over the first RF chain and the second RFchain.

[0309] 62. The method of embodiment 61, wherein the first RF chainand the second RF chain are calibrated.

[0310] 63. The method of any one of embodiments 56-62, wherein thetransmitting Nt known training sequences sweeping the transmitantenna elements comprises the initiator station having two RFchains and having transmit antenna elements, transmitting Nt knowntraining sequences sweeping the transmit antenna elements, whereina number of the transmit antenna elements is larger than a numberof RF chains.

[0311] 64. The method of any one of embodiments 56-63, wherein thenumber of transmit antenna elements is at least six times largerthan a number of RF chains.

[0312] 65. The method of any one of embodiments 56-64 furthercomprising the responder transmitting the estimated two channels tothe initiator.

[0313] 66. A method for beamforming training for spatialmultiplexing, the method comprising: [0314] an initiator stationhaving two or more radio frequency (RF) chains and having transmitantenna elements, the initiator station transmitting Nt knownsequences sweeping the transmit antenna elements; [0315] aresponder station having a first and a second RF chain and havingreceive antenna elements, and having a first and a second receivebeamforming weight, the responder station receiving the first RFchain and receiving the second RF chain; and [0316] the responderestimating at least two channels by comparing the received first RFchain and the second RF chain to the known Nt trainingsequences.

[0317] 67. The method of embodiment 66, wherein the first RF chainis a combination of signals received from the receive antennaelements weighted by the first receive beamforming weights.

[0318] 68. The method of embodiment 66 or 67, wherein the second RFchain is the combination of signals received from the receiveantenna elements weighted by the second receive beamformingweights.

[0319] 69. A method for performing beamforming, the methodcomprising: [0320] a first communication device transmitting afirst plurality of beamforming training frames to a secondcommunication device using a first beamforming vector; [0321] thefirst communication device receiving from the second communicationdevice a second beamforming weight vector; and [0322] the firstcommunication device transmitting a second plurality of beamformingtraining frames to the second communication device using the secondbeamforming vector.

[0323] 70. The method of embodiment 69, wherein the firstcommunication device transmits the first plurality of beamformingtraining frames to the second communication device using a firstbeamforming weight vector.

[0324] 71. The method of embodiment 69 or 70, wherein the firstcommunication device transmits a first portion of the firstplurality of beamforming training frames using a first group ofantenna and a first portion of the beamforming weights.

[0325] 72. The method of any one of embodiments 69-71, wherein thefirst communication device transmits a second portion of the firstplurality of beamforming training frames using a second group ofantenna and a second portion of the beamforming weights.

[0326] 73. A base station configured to perform any one ofembodiments 53-72.

[0327] 74. A base station configured to perform any portion ofembodiments 53-72.

[0328] 75. An integrated circuit configured to perform any one ofembodiments 53-72.

[0329] 76. An integrated circuit configured to perform any portionof embodiments 53-72.

[0330] 77. A station (STA) configured to perform any one ofembodiments 53-72.

[0331] 78. A station (STA) configured to perform any portion ofembodiments 53-72.

[0332] 79. An access point (AP) configured to perform any one ofembodiments 53-72.

[0333] 80. An access point (AP) configured to perform any portionof embodiments 53-72.

[0334] 81. A wireless transmit/receive unit (WTRU) configured toperform any one of embodiments 53-72.

[0335] 82. A wireless transmit/receive unit (WTRU) configured toperform any portion of embodiments 53-72.

[0336] Although features and elements are described above inparticular combinations, one of ordinary skill in the art willappreciate that each feature or element can be used alone or in anycombination with the other features and elements. In addition, themethods described herein may be implemented in a computer program,software, or firmware incorporated in a computer-readable mediumfor execution by a computer or processor. Examples ofcomputer-readable media include electronic signals (transmittedover wired or wireless connections) and computer-readable storagemedia. Examples of computer-readable storage media include, but arenot 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, anddigital versatile disks (DVDs). A processor in association withsoftware may be used to implement a radio frequency transceiver foruse in a WTRU, UE, terminal, base station, RNC, or any hostcomputer.

* * * * *