Media access control for MIMO wireless network

[0001] The present application is related to the subject matter of U.S. Pat. No. 6,377,631, issued on Apr. 23, 2002, and entitled “TRANSMITTER INCORPORATING SPATIO-TEMPORAL PROCESSING,” the contents of which are incorporated by reference herein for all purposes in their entirety.

[0002] The present invention relates generally to communications and more particularly to systems and methods for wireless communications.

[0003] As the Internet continues its development and as workers and consumers increasingly rely on data networking to assist their day-to-day tasks, a need arises to extend network connectivity to locations where there is no convenient connection to a wired infrastructure. Workers desire to send and receive email and access the Internet and their corporate intranet even when they are away from their workstation. Consumers wish to establish home networks without costly and cumbersome wiring. Accordingly, wireless communication standards have evolved including the IEEE 802.11 family.

[0004] The current IEEE 802.11a standard allows for wireless communicates at speeds between 6 Mbps and 54 Mbps. It is desirable to further increase these speeds to accommodate delivery of multimedia wireless services and facilitate outdoor wireless bridging between indoor networks. It is also desirable to accommodate the increased data rates by increasing spectral efficiency rather than by increasing bandwidth.

[0005] One known way of increasing spectral efficiency is the use of MIMO (Multiple Input Multiple Output) processing techniques. MIMO techniques take advantage of multiple antennas (or multiple polarizations of the same antenna) at the transmitter and receiver to access multiple channel inputs and channel outputs and thereby define multiple spatial subchannels that occupy the same bandwidth but nonetheless are capable of carrying independent data streams. The delineation of the multiple spatial subchannels may involve weighting of the antenna inputs at the transmitter end and/or weighting of the antenna outputs at the receiver end. For further information on MIMO techniques, see U.S. Pat. No. 6,377,631.

[0006] It is desirable to apply MIMO techniques to IEEE 802.11 systems to increase data carrying capacity but there are obstacles to overcome. The 802.11 standards do not specify MIMO transmission techniques. It is desirable, and probably commercially essential, to construct wireless communication devices that cannot only interoperate with MIMO-capable devices but also with standards-defined devices that are not MIMO-capable. It is also desirable to modify the media access control (MAC) layer of standards such as the 802.11 standard to facilitate MIMO communications and allow 802.11 networks and other networks to benefit from increased capacity.

[0007] By virtue of one embodiment of the present invention, the MAC layer of a wireless communication standard is modified to facilitate MIMO (multiple input multiple output) processing and thereby increase capacity and/or spectral efficiency. In one implementation, the IEEE 802.11a standard is modified in this way. A contention free period (CFP) is divided between MIMO communications and non-MIMO communications. An access point may maintain a list of nodes operating in MIMO mode and nodes operating in non-MIMO mode.

[0008] A first aspect of the present invention provides a method for operating an access node in a wireless communication system having nodes operating in a MIMO mode and nodes operating in a non-MIMO mode. The method includes: maintaining a list of nodes operating in a MIMO mode and nodes operating in a non-MIMO mode and, within a designated contention free period, polling MIMO-mode nodes and non-MIMO-mode nodes for transmissions in corresponding non-overlapping subperiods of the contention free period.

[0009] A second aspect of the present invention provides a method for a method of operating a MIMO-capable subscriber node. The method includes: during a contention period, requesting an access point for permission to switch between a MIMO mode and a non-MIMO mode, and upon receiving permission, switch between the MIMO mode and the non-MIMO mode.

[0010] A third aspect of the present invention provides apparatus for operating an access node in a wireless communication system having nodes operating in a MIMO mode and nodes operating in a non-MIMO mode. The apparatus includes: a MIMO-capable physical layer transceiver and a MAC layer block that maintains a list of nodes operating in a MIMO mode and nodes operating in a non-MIMO mode, and within a designated contention free period, polls MIMO-mode nodes and non-MIMO-mode nodes for transmissions in corresponding non-overlapping subperiods of the contention free period.

[0011] A fourth aspect of the present invention provides apparatus for operating a MIMO-capable subscriber node. The method includes a MIMO-capable physical layer transceiver and a MAC layer block that, during a contention period, requests an access point for permission to switch between a MIMO mode and a non-MIMO mode upon receiving permission, switches between the MIMO mode and the non-MIMO mode.

[0012] Further understanding of the nature and advantages of the inventions herein may be realized by reference to the remaining portions of the specification and the attached drawings.

[0021] Although having very broad applicability, the present invention will be described with reference to a representative network environment, a wireless communication network based on the IEEE 802.11 standard, and in one particular implementation, the IEEE 802.11a standard. The IEEE 802.11g standard is another example of a wireless communication standard to which embodiments of the present invention may be applied. Familiarity with the technical details of these standards will be assumed in the discussion that follows. Relevant descriptive materials regarding the IEEE 802.11 standard may be found in the following documents:

[0022] Information technology—Telecommunications and information exchange between systems—Local and metropolitan area networks specific requirements—Part 11: Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) Specifications (1999).

[0023] Information technology—Telecommunications and information exchange between systems—Local and metropolitan area networks specific requirements—Part 11: Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) Specifications (1999): High Speed Physical Layer in the 5 GHz band.

[0024] The contents of these standards documents are herein incorporated by reference for all purposes in their entirety.

[0025] FIG. 1 depicts a wireless network 100 suitable for implementing one embodiment of the present invention. Included within wireless network 100 are an access point 102 and numerous subscriber units 104. In one application, subscriber units 104 are themselves wireless bridges to other wireless networks. For example, network 100 may be an outdoor wireless backbone interconnecting a number of in-building wireless networks.

[0026] Access point 102 is equipped with two antennas 106. Alternatively, the two depicted antennas 106 actually signify the use of two polarizations of the same antenna. Some, but not all, of subscriber units 104 are also equipped with two antennas 108 or able to exploit dual polarizations. These subscriber units are capable of MIMO communications with access point 102. Other subscriber units have but a single antenna 110 and are therefore not capable of MIMO communications. The MIMO-capable subscriber units may operate in MIMO mode or conventional “SISO” (Single Input Single Output) mode depending on channel conditions as will be explained.

[0027] In one implementation, access point 102 and subscriber units 104 communicate in accordance with the IEEE 802.11a standard. Operation is typically in accordance with so-called “infrastructure” mode where communication is either to or from access point 102 rather than between subscriber units 104. Furthermore, access point 102 and the MIMO-capable ones of subscriber units 104 incorporate MIMO-related extensions as described below.

[0028] Each of access points 102 and subscriber units 104 incorporates both a transmitter and a receiver. FIG. 2 depicts elements of a transmitter 200 as would be used in MIMO operation in accordance with one embodiment of the present invention. A MAC layer processor 202 performs functions relating to coordinating access to the wireless transmission medium in accordance with the IEEE 802.11a standard as extended to accommodate MIMO processing. MAC layer processor 202 also acts as an interface between physical layer elements and higher layer protocols that sink and source data. The exact functions of MAC layer processor 202 will vary depending on whether transmitter 200 is operating in an access point 102 or subscriber unit 104.

[0029] MAC layer processor 202 may be implemented in any suitable manner, e.g., as an ASIC, FPGA, a programmable processor operating suitable software, etc. If MAC layer processing is based on software, the software to implement MAC layer functions may be stored on a computer-readable storage medium 204. Computer-readable storage medium 204 may represent on-chip or off-chip ROM, RAM, etc. Other storage media such as floppy disks, or CD-ROMs, may be used to store software useful in implementing embodiments of the present invention.

[0030] When operating in MIMO mode, transmitter 202 transmits at twice the data rate of conventional 802.11a units given the same selection of constellation and coding rate. Accordingly, the MAC layer packet is twice as large but divided into 2 for transmission via 2 spatial subchannels. Also, the length field of the physical layer header (PLCP header in 802.11 terminology) is set to ½ of the length of the MAC layer packet. If half the MAC layer packet length is not an integer, the length field is set to the next largest integer.

[0031] After division of the MAC layer packet into two, further physical layer processing is duplicated for each spatial subchannel. Each packet half is given its own PLCP preamble that includes the symbols used for training and synchronization in accordance with the 802.11a standard. For each spatial subchannel, there is a scramble/encode/interleave block 206 that performs error correction coding, scrambling, and interleaving in accordance with the 802.11a standard. A modulator 208 maps bits to particular constellation points corresponding to each subcarrier position within an OFDM (Orthogonal Frequency Division Multiplexing) symbol. Pilot and zero subcarriers are inserted as required by the standard. An IFFT block 210 converts sets of subcarrier values to time domain samples representing an OFDM symbol using the inverse Fast Fourier transform. Preamble training symbols are inserted as required. A digital to analog converter 212 converts the time domain sample stream to an analog signal.

[0032] Further analog processing at one or more intermediate frequencies (IFs) and at the transmission radio frequency (RF) is also duplicated for each subchannel. In one embodiment, each subchannel RF output is connected to a different polarization of the same antenna. Alternatively, the RF outputs can drive different antennas.

[0033] In more complex, MIMO processing schemes, a 2×2 matrix weighting processor is used to convert the time domain sample streams output of each spatial subchannel to the signals driving each antenna or antenna polarization. The weighting matrix may be derived from a channel estimate, a channel estimate enhanced with SINR (signal to interference plus noise ratio) information, etc. Also, the weighting matrix may be adjusted prior to the IFFT and adjusted for each subcarrier.

[0034] When operating in a SISO mode, transmitter 200 may make use of only a single chain of blocks 206, 208, 210, and 212 and a single IF/RF chain. Only a single antenna polarization is used then.

[0035] FIG. 3 depicts a receiver 300 as would be used in one access point 102 or one of the MIMO-capable ones of subscriber units 104. Individual RF/IF chains (not shown) are connected to each antenna (or antenna polarization) to recover analog signals suitable for conversion to digital samples from the RF signals received over the airwaves. There is an analog to digital converter 302 for each analog signal to recover digital samples. There are also two parallel FFT blocks 304 to perform time to frequency conversion and recover the subcarrier values for each OFDM symbol. Components to perform time and frequency synchronization in accordance with the 802.11a standard are omitted for simplicity of explanation.

[0036] The subcarrier values corresponding to each antenna or polarization are input to a beamformer 306. Beamformer 306 recovers the transmitted spatial subchannel data streams by applying a weighting matrix to the beamformer inputs. The weighting matrix may be based on a matrix channel estimate and possibly also on (signal to interference plus noise ratio) SINR. There are many suitable ways of implementing the weighting matrix but this type of detail is not relevant to a discussion of the present invention. A pilot tone correction block 308 corrects phase offset among subcarriers within each OFDM symbol in accordance with the 802.11a standard. For each spatial subchannel, a decode block 310 performs error correction decoding, descrambling, and deinterleaving to reverse the corresponding processes at the transmit end. The outputs of decode block 310 are recombined into MAC layer packets by MAC layer processor 202 which is shared with transmitter 200. FIG. 3 is depicted with reference to MIMO operation. For SISO operation, a single conventional 802.11a receiver chain is used without any beamforming processing.

[0037] FIG. 4 depicts the operation of wireless network 100 in the time domain according to one embodiment of the present invention. It will be appreciated that the operation of wireless network 100 shares features in common with conventional 802.11 “infrastructure mode” networks. Transmission time is divided between a contention period (CP) and a contention free period (CFP). During the CP devices transmit when they do not hear other transmissions. During the CFP, access point 102 controls access to the transmission medium and polls individual subscriber units 104 to transmit data that they may have. Downstream data may be sent along with polling information.

[0038] In 802.11 operation, each CFP begins with a special beacon frame that defines the beginning of the CFP and indicates its duration. According to one embodiment of the present invention, this beacon is followed by a special MIMO beacon that is transmitted in MIMO mode whereas the first beacon is transmitted in SISO mode. MIMO-capable subscriber units that are in SISO mode listen to the first beacon to learn about the length of the CFP but temporarily switch to MIMO mode to monitor the MIMO beacon and make measurements on the MIMO beacon preamble to estimate MIMO channel conditions. These measurements help determine whether they should in fact switch from SISO mode to MIMO mode. In one embodiment, cross-polarization discrimination (XPD) between the two spatial subchannels is measured and if XPD is sufficiently high, the subscriber unit will switch over to take advantage of the greater capacity available in MIMO mode. In an alternative embodiment, error vector magnitude (EVM) is determined and if it is sufficiently high, the subscriber unit will switch over. The EVM measurement takes into effect all of the wireless medium channel impairments, noise, interference, and cross-polarization discrimination.

[0039] MIMO-capable subscriber units in MIMO mode will receive all relevant beacon information from the MIMO beacon when in MIMO mode. Non-MIMO-capable subscriber units will not be capable of receiving the MIMO beacon and will typically ignore it as directed by the SISO beacon parameters. A short time period TS/M separates the two beacons.

[0040] Access point 102 maintains of lists of subscriber units operating in MIMO node and subscriber units operating in SISO mode (including non-MIMO-capable subscriber units.) The CFP is divided into two subperiods. In a first subperiod, access point 102 communicates with the subscriber units that are operating in MIMO mode. In a second subperiod, access point 102 communicates with the subscriber units that are operating in SISO mode. A time TM/S separates the subperiods. Subscriber units 104 communicate data in response to a polling request from access point 102 as defined by the IEEE 802.11 standard and/or they receive downstream data from access point 102. The division of the CFP into subperiods reduces the number of time-consuming mode changes for access point 102 to the minimum. If there is time left in the CFP, access point 102 can repeat the subperiods to fill up the available time. Upon completion of the CFP, access point 102 transmits the 802.11-defined CF-End message in both SISO mode and MIMO mode so that all units know that the contention period (CP) will begin.

[0041] In one embodiment, all CP transmissions occur in SISO mode. According to the present invention, these may include specially modified messages for affiliating to the network as a MIMO-capable node and for requesting and acknowledging mode shifts. In particular, embodiments of the present invention may take advantage of Association, Reassociation, Disassociation Request, and Response Frames as specified by 802.11. These frames may be augmented with a special MIMO information element. In one implementation, this MIMO information element has a length of 1 octet. The least significant bit indicates SISO operation or MIMO operation while the remaining bits can be used for version identification.

[0042] When a MIMO-capable subscriber unit starts up, it transmits an Association Request to access point 102 in accordance with 802.11a. However, this Association Request includes a MIMO information element that specifies MIMO mode operation in accordance with the present invention. Access point 102 replies with an Association Response message that also includes the MIMO information element specifying MIMO mode operation. The subscriber unit then switches to MIMO mode and access point 102 adds the subscriber unit to the list of MIMO-capable subscriber units. If the subscriber unit joins a network where the access point is not MIMO-capable, the received Association Response will lack the MIMO information element and the subscriber unit will therefore remain in SISO mode. This preserves backward compatibility with non-MIMO equipment. FIG. 5 depicts a state transition diagram showing start up for a MIMO-capable subscriber unit.

[0043] MIMO-capable subscriber units preferably switch between MIMO-mode and SISO-mode depending on channel conditions. In one embodiment, when in MIMO mode, the subscriber unit measures an indicator of the isolation of the two spatial subchannels. In embodiments where antenna polarizations are utilized to establish the spatial subchannels, cross-polarization discrimination (XPD) may be used. These measurements are made on MIMO transmissions from access point 102 while the subscriber unit is in MIMO mode. In one implementation, the measurements are made on a special preamble incorporated into the MIMO transmissions. When the subscriber unit is in MIMO mode, it may make this measurement on any downstream transmission. When the subscriber unit is otherwise in SISO mode, it will temporarily shift to MIMO mode to make this measurement on the MIMO beacon.

[0044] Before switching between MIMO mode and SISO mode, a subscriber unit performs a handshake with access point 102 via a Reassociation Request/Response exchange during the CP. As earlier indicated, all transmissions during the CP are SISO transmissions regardless of the current mode of the subscriber unit. The Reassociation Request and Reassociation Response messages will include the above-mentioned MIMO information element with the appropriate bit set to reflect the new mode of the subscriber unit.

[0045] FIG. 6 depicts the state transition diagram for a subscriber unit that is initially in MIMO mode and may change to SISO mode. As long as the XPD is sufficiently high, the subscriber unit remains in MIMO mode. If the XPD test fails, the subscriber unit waits for CP and switches to SISO mode and sends a Reassociation Request indicating SISO mode is now set. If this Reassociation Request is confirmed with a Reassociation Response indicating that SISO mode is now set, the subscriber unit remains in SISO mode. Otherwise, the subscriber unit returns to MIMO mode. Also, in addition to confirming with a Reassociation Request, access point 102 moves the subscriber unit from the MIMO mode list to the SISO mode list.

[0046] FIG. 7 depicts the state transition diagram for a subscriber unit that is initially in SISO mode and may change to MIMO mode. The subscriber unit shifts to MIMO mode at the scheduled time of the MIMO beacon and then shifts back to SISO mode after monitoring the beacon and measuring XPD. If the XPD test fails, the subscriber unit then remains in SISO mode. If the XPD test passes, the subscriber unit sends its Association Request indicating a shift to MIMO mode. If a confirmation is received, the subscriber unit shifts to MIMO mode. Access point 102 also moves the entry for the subscriber unit from the SISO list to the MIMO list. If the Association Request is not confirmed, the subscriber unit remains in SISO mode.

[0047] FIG. 8 depicts the state transition diagram for access point 102. During the CP, and also during the CFP when polling the SISO subscriber units, access point 102 remains in SISO mode. Access point 102 shifts to MIMO mode to send a MIMO beacon whenever the SISO beacon has just been sent and there is at least one listed subscriber unit in MIMO-mode. Access point 102 also shifts from SISO mode to MIMO mode during the CFP after the SISO subscriber units have been polled if there is time to again poll the MIMO subscriber units. Access point 102 remains in MIMO mode during the CFP while polling MIMO-mode subscriber units and then returns to SISO mode after completion of polling of the MIMO-mode subscriber units or at the end of the CFP.

[0048] It will be seen that MIMO capability has been brought to an IEEE 802.11 network without sacrificing backward compatibility. Also, mechanisms for shifting to and from MIMO mode depending on channel conditions have been provided.

[0049] It is understood that the examples and embodiments that are described herein are for illustrative purposes only and that various modifications and changes in light thereof will be suggested to persons skilled in the art and are to be included within the spirit and purview of this application and scope of the appended claims and their full scope of equivalents. For example, it will be appreciated that although the present invention has been explained with reference to an IEEE 802.11 wireless network, it will find application in other types of network as well.