ES2357849T3 - OPEN LOOP SPEED CONTROL FOR A TDD COMMUNICATIONS SYSTEM. - Google Patents

Abstract

A method for making a speed selection, comprising: estimating (114) the channel quality of a first communications link based on a transmission received through the first communications link; estimate (116) the channel quality of a second communications link based on the estimated channel quality of the first communications link and an asymmetric parameter indicative of a difference in the channel qualities of the first and second communications link, where the difference in channel qualities is the difference between an SNR for the first communications link and an SNR for the second communications link; and select (118) at least one speed for a data transmission through the second communications link based on the estimated channel quality of the second communications link.

Description

BACKGROUND

I. Field

The present invention relates generally to communications and, more specifically, to speed control for a communications system.

II. Background

In a wireless communications system, a transmitting station processes traffic data for one or more data streams, generates one or more modulated signals and transmits the modulated signal (s) through a wireless channel to a receiving station The modulated signal (s) may undergo several harmful channel conditions such as, for example, signal fading, multipath and interference effects. The receiving station receives the modulated signal (s) and processes one or more received signals to retrieve traffic data.

Speed control refers to the process for controlling the speed for each data stream sent by the transmitting station to the receiving station. A "speed" may be associated with a particular data transfer rate, a particular coding scheme or a code rate, a particular modulation scheme, etc., to be used for a data stream. The speed control tries to select a speed that is as high as possible while at the same time adjusting to a target packet error rate (PER) for each data stream based on the channel conditions so that the data stream can be received reliably.

The speed control is normally carried out in the manner of a closed loop. For closed loop speed control, the transmitting station normally transmits a pilot signal that is used by the receiving station to estimate the channel quality, which can be quantified by a signal to noise ratio plus interference (SNR). Then, the receiving station selects one or more speeds based on the estimated channel quality and sends the selected speed (s) to the transmitting station. The transmitting station processes the traffic data based on the selected speed (s). GB 2 350 753 A, EP 0 868 034 A2 and US 2001/0048675 A1 show examples for such closed loop systems.

Closed loop speed control is generally effective but may not always be available. In addition, closed loop speed control has some disadvantages. First, the transmitting station uses system resources to transmit the pilot signal to the receiving station, and the receiving station uses system resources to return the selected speed (s). Secondly, additional delays are usually generated at the transmitting station to transmit the pilot signal and at the receiving station to return the selected speed (s).

Therefore, there is a need in the art for technologies to carry out speed control when closed loop speed control is not available or when good performance is desired using less system resources, for example supplementary transmission data to communicate channel quality, and with less delay.

SUMMARY

This need is met by the method according to independent claim 1 and by the apparatus according to independent claim 19. This document describes techniques for carrying out an open-loop speed control in a time division duplex communications system ( TDD). According to an embodiment of the invention, an apparatus is described that includes a processor and a controller. The processor estimates the channel quality of a first communications link based on a transmission (for example, a pilot signal transmission) received through the first link. The controller estimates the channel quality of a second communications link based on the estimated channel quality of the first link and an asymmetric parameter indicative of a difference in the channel qualities of the first and second links.

According to another embodiment, an apparatus is described that includes a processor and a controller. The processor obtains a first set of SNR estimates for a first set of transmission channels in a first link based on a transmission received through the first link. The controller obtains a second set of SNR estimates for a second set of transmission channels on a second link based on the first set of SNR estimates and an asymmetric parameter. The controller also selects a set of speeds for the second set of transmission channels based on the second set of SNR estimates.

According to another additional embodiment, an apparatus is described that includes a processor and a controller. The processor obtains at least one SNR estimate for a first link based on a transmission received through the first link. The controller obtains at least one SNR estimate for a second link based on the at least one SNR estimate for the first link. The controller also selects at least one speed for at least one data stream based on the at least one SNR estimate for the second link and adjusts the transmission of the at least one data stream based on feedback received for packets in the at least a flow of data

According to another additional embodiment, an apparatus is described that includes a controller and a first and a second processor. The first processor determines the SNRs received for a first set of subbands of a first transmission channel in a first link based on a pilot signal transmission received through the first link. The controller selects at least one subband from a second set of subbands of a second transmission channel on a second link based on the SNRs received for the first transmission channel. The second processor processes data for transmission in the at least one subband of the second transmission channel.

Various other aspects and embodiments of the invention are described in greater detail below.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 shows a process carried out by a transmitting station for open loop speed control according to one or more embodiments.

FIG. 2 shows a process for carrying out an open loop speed control in a TDD system according to one or more embodiments.

FIG. 3 shows a block diagram of a transmitting station and a receiving station according to one or more embodiments

DETAILED DESCRIPTION

The expression "by way of example" is used in this document with the meaning of "serving as an example, instance or illustration." Any embodiment described in this document as "by way of example" should not necessarily be regarded as preferred or advantageous over Other realizations

The open loop speed control techniques described in this document can be used for a single input and single output system (SISO), for a single input and multiple output system (SIMO), for a multiple input and single output system (MISO) and for a system of multiple inputs and multiple outputs (MIMO). A MIMO system uses multiple (T) transmission antennas at a transmitting station and multiple (R) receiving antennas at a receiving station for data transmission. A MIMO channel formed by the T transmission antennas and the R reception antennas can be broken down into S space channels, where S ≤ min {T, R}. Much of the following description is for a MIMO system.

Open loop speed control techniques can also be used for single carrier and multi carrier systems. Multiple carriers can be obtained with orthogonal frequency division multiplexing (OFDM) techniques or with some other multi-carrier modulation technique. OFDM divides the overall bandwidth of the system into multiple (K) orthogonal frequency subbands, which are also referred to as tones, subcarriers, bins or frequency channels. With OFDM, each subband is associated with a respective subcarrier that can be modulated with data. For a MIMO system that uses OFDM, which is referred to as a MIMO-OFDM system, there are S space channels available for each of the K subbands. S broadband space channels can be formed with the space channels of the K subbands.

For reasons of clarity, the following description assumes that a data stream is sent on each transmission channel. A transmission channel may be a space channel in a MIMO system, an orthogonal space channel (or own mode) in a MIMO system that transmits in its own modes, a broadband space channel in a MIMO-OFDM system, a proprietary mode of broadband in a MIMO-OFDM system that transmits in its own modes, etc. A transmission channel can also be referred to as a parallel channel, a data channel, a traffic channel, a physical channel, or as a term of another terminology.

For a calibrated MIMO TDD system, it can be assumed that the responses of the wireless channel between two stations are reciprocal to each other. Therefore, if the matrix H represents the channel response from a station A to a station B, then a reciprocal channel implies that the channel response from station B to station A is given by HT, where HT denotes the transpose of H. For a reciprocal channel it can be assumed that the changes in the massive loss of trajectory, overshadowing, multipath, signal fade, etc., are identical for the two links of the wireless channel. This reciprocity can be exploited for open loop speed control.

The link from station A to station B can be denoted as link (A, B), and the link from station B to station A can be denoted as link (B, A). For station A, link (A, B) is the transmission link and link (B, A) is the reception link. For station B, link (B, A) is the transmission link and link (A, B) is the reception link. The link (A, B) and the link (B, A) are considered to be symmetric if the SNR observed in the

5 station A for a transmission from station B is the same as the SNR observed at station B for a transmission from station A, assuming that the channel conditions have not changed in the time interval of interest. Symmetric links can be achieved for two stations A and B, for example, if the characteristics of the transmitting unit and the receiving unit are the same for these two stations.

In practice, symmetric links can be difficult to achieve due to several factors such as the diversity of

10 manufacturing, component tolerance, etc. In addition, different stations can be manufactured with different capacities, for example, different levels of transmission power, different receiver noise factors, different matrix dimensions, etc. As a result, the links for stations A and B may not be symmetrical, as illustrated below.

Table 1 shows example capabilities for stations A and B.

15 Table 1

 Potency of transmission

Receiver noise factor Number of receiving antennas

Station A

17 dBm 6 dB 4

Station B

14 dBm 10 dB 2

For the example shown in Table 1, the received signal level (RSL) observed at station B when station A transmits at its maximum power in a lossless channel can be calculated as:

image 1

The RSL observed at station A when station B transmits at maximum power on a lossless channel can be calculated as:

image2

An asymmetric ASYM parameter (t, r) can be defined as the difference between (1) the SNR observed at station r when station t transmits at a known power level in a known channel and (2) the SNR observed at the station t when station r transmits. For the example described above, the asymmetric parameters for stations A and B can be calculated as:

image3

image4

When the links are symmetric, the asymmetric parameters can be given as:

image4

When the links are asymmetric, as is the case in the example shown in Table 1, the asymmetric parameters are non-zero and are determined by the difference in SNRs observed by the two stations for transmissions from the other station. A positive value for the ASYM parameter (t, r) denotes that the SNR observed by station r from station t is greater than the SNR observed by station t from station r. Conversely, a negative value for the ASYM parameter (t, r) denotes that the SNR observed by station r from station t is smaller than the SNR observed by station t from station r.

If the asymmetric parameters are known, then each station can infer directly from its receiver the speeds it can use to transmit data to the other station. For example, if station A transmits pilot signals and / or data to station B, then station B can estimate the SNR for each transmission channel on the link (A, B) based on the pilot signals and / or the signals. data received from station A. Then, station B can estimate the SNR for each transmission channel on the link (B, A) as follows:

image4

where

(A, B) is a measurement of SNR for the transmission channel m on the link (A, B); Y

 m

ˆ (B, A) is an estimate of SNR for the transmission channel m on the link (B, A).

m

Station B can obtain SNR estimates for the transmission channels on the link (A, B) based on the pilot signals and / or the received data, as described below. Then, station B can obtain SNR estimates for the transmission channels on the link (B, A) based on the SNR estimates for the link (A, B) and the ASYM parameter (A, B), as and as shown in equation (6). Then, station B can use SNR estimates for link (B, A) to select suitable speeds for data streams sent through the transmission channels on link (B, A) to station A.

Also, if station B transmits pilot signals and / or data to station A, then station A can obtain an SNR estimate for each transmission channel on the link (B, A) based on the pilot signals and / or data received from station B. Then, station A can obtain an SNR estimate for each transmission channel on link (A, B) as follows:

image4

where

(B, A) is an SNR measurement for the transmission channel m on the link (B, A), and

 m

ˆ (A, B) is an estimate of SNR for the transmission channel m on the link (A, B).

m

Station A can use ˆ (A, B) to select a suitable speed for each transmission channel m in the

m

link (A, B).

The ASYM parameter can be determined for each link in several ways. In one embodiment, stations A and B exchange their capabilities (for example, transmission power, noise factor and number of antennas) by signaling. Each station can then calculate the ASYM parameter for its reception link based on the capabilities of both stations. In another embodiment, each station measures the SNR received for its reception link based on a pilot signal transmitted at a power level known to the other station and sends the received SNR to the other station. Each station can then calculate the ASYM parameter based on the SNRs received for the two links and the power levels of the pilot signals sent on the two links. In another additional embodiment, the ASYM parameter is determined based on receiver feedback without any initial message exchange. The ASYM parameter can be initialized as ASYM (A, B) = 0. When station B transmits to station A, the station

A calculates the estimate of SNR as ˆ (B, A) =  (A, B) - ASYM (A, B), where ASYM (A, B) is initially zero. The

mm

Feedback from station A is used to adjust the ASYM parameter in order to achieve the desired packet error rate. The feedback from station A can be ACK / NAK provided by the MAC layer or other feedback (for example, real-time control protocol (RTCP) frames).

In many cases, the ASYM parameter for each link is a fixed or static value. For these cases, the ASYM parameter can be determined once for each link and used for all subsequent data transmissions sent through that link. However, there may be situations in which the ASYM parameter varies. For example, the dynamic range of the stations may be different, and nonlinearities may modify the value of the ASYM parameter depending on the actual signal levels received. In addition, the ambient temperature may affect the receiver noise factor. As another example, different characteristics of phase noise in the transmitter and receiver chains can also influence the ASYM parameter depending on the SNR received. In these cases, changes and / or errors in the ASYM parameter can be represented by an external loop, as described later.

FIG. 1 shows a process 100 carried out by the transmitting station for an open loop speed control according to one or more embodiments. Initially, information can be exchanged with the receiving station to determine the ASYM parameter (block 112). Block 112 can also be omitted if the ASYM parameter is determined without any initial message exchange, as described above. The channel quality of a first transmission channel in a first communication link is then estimated based on the transmission of pilot and / or data signals received through the first communication link (block 114). Then the channel quality of a second transmission channel in a second communication link is estimated based on the estimated channel quality for the first transmission channel and the ASYM parameter (block 116). A speed for the second transmission channel is selected based on the estimated channel quality for the second transmission channel (block 118). The speed, the ASYM parameter and / or the speed adjustment parameter can be adjusted depending on the performance of a data transmission sent through the second communication link (block 120).

FIG. 2 shows a process 200 for carrying out an open loop speed control in a TDD system according to one or more embodiments. Initially, stations A and B exchange relevant information so that each station can determine the ASYM parameter for its reception link (blocks 210 and 212). Blocks 210 and 212 may be carried out once (for example, during system access or registration), periodically or when necessary.

Then, for data transmission from station A to station B, station B transmits a pilot signal on link (B, A) (block 214). Station A receives the pilot signal and obtains SNR estimates for the transmission channels on the link (B, A), as described later (block 216). Then, station A obtains SNR estimates for the transmission channels on the link (A, B) based on the SNR estimates for the link (B, A) and the ASYM parameter (block 218). Station A selects speeds for the transmission channels on the link (A, B) based on SNR estimates for those transmission channels (block 220). Then, station A transmits data streams at the selected speeds through the transmission channels on the link (A, B) (block 222).

Station B receives the data streams from station A and decodes the packets sent in those data streams (block 224). Station B can send receipt confirmations (ACK) for correctly decoded packets (or valid packets) and / or negative receipt confirmations (NAK) for failed decoded packets (packets removed) (block 226). Station A receives the ACK / NAK feedback and adjusts the ASYM parameter, the speeds and / or the speed adjustment parameters based on the ACK / NAK feedback (block 228). Station A adjusts the transmission of the data streams based on the settings determined in block 228.

FIG. 3 shows a block diagram of station A 310 and station B 350 according to one or more embodiments. For downlink transmission, station A is an access point, station B is a user terminal, link (A, B) is the downlink or direct link, and link (B, A) is the uplink or reverse link. For uplink transmission, station A is a user terminal, station B is an access point, link (A, B) is the uplink and link (B, A) is the downlink. For a peer-to-peer transmission, stations A and B can be two user terminals.

At station A, a speed selector / controller 334 selects a speed for each of the M data streams to be sent to station B, where M  1. The M data streams can be sent at the same rate or different speeds, depending on the way data flows are transmitted and / or channel conditions. A transmission data processor (TX) 320 receives traffic data, processes (for example, encodes, interlaces and correlates symbols) the traffic data based on the selected M speeds, and generates M data symbol streams. As used herein, a data symbol is a modulation symbol for traffic data, a pilot symbol is a modulation symbol for a pilot signal (which is data known a priori by both the transmitting station and the receiver), a modulation symbol is a complex value for a point in a constellation of signals for a modulation scheme (for example, M – PSK or M – QAM), and a symbol is a complex value. A TX 322 space processor multiplexes the M data symbol flows with pilot symbols, performs spatial processing on the data and the pilot symbols if applicable and, as described below, provides T streams of transmission symbols for T transmission antennas. A transmitter unit (TMTR) 324 processes the T streams of transmission symbols (for example, of OFDM) and generates T modulated signals, which are transmitted through the link (A, B) to the station B. The link (A , B) distorts the modulated signals with a channel response and further degrades the modulated signals with additive Gaussian white noise (AWGN) and possibly with interference from other stations.

At station B, a receiver unit (RCVR) 360 receives the signals transmitted through R reception antennas and processes (for example, conditions and digitizes) R signals received to generate R sample flows. The receiving unit 360 further processes the R sample flows (eg, of OFDM), provides pilot symbols received to a channel / processor estimator 362 and provides data symbols received to a spatial receiving processor (RX) 364. The estimator channel 362 estimates the MIMO channel response for link (A, B) and provides channel estimates to the RX 364 space processor. The RX 364 space processor performs spatial receiver processing on the received data symbols based on channel estimates and provides M flows of detected data symbols, which are estimates of the M flows of data symbols sent by station A. An RX 368 data processor processes (for example, de-maps symbols, deinterlaces and decodes) the M flows of data symbols detected based on the selected M speeds and provides decoded data, which are an estimate of the d Traffic atos sent by station A. The RX 368 data processor also provides the status of each packet received (for example, deleted or valid).

Controllers 340 and 390 control the operation of several processing units at stations A and B, respectively. Memory units 342 and 392 store data and program codes used by controllers 340 and 390, respectively.

Open loop speed control techniques do not require that the receiving station B send explicit feedback information to the transmitting station A for speed selection specifically. The term "open loop" in the context of open loop speed control refers to the absence of such explicit feedback information, which may include, for example, one speed per data stream, one SNR per data stream, one SNR per antenna, etc. On the contrary, a closed loop speed control scheme normally requires the receiving station to determine feedback information based on the channel conditions observed at the receiving station and to send this information to the transmitting station, which then uses the information for speed selection for the transmission of data to the receiving station.

Open loop speed control techniques use information that is available for normal system operation. For example, open loop speed control techniques can use (1) received SNRs obtained by the transmitting station A based on the pilot signals and / or the data received from the receiving station B, (2) the ACKs and / or the NAKs received from station B for packets sent by station A, and (3) possibly other information. A media access control (MAC) layer that resides above a physical layer on station B can send an ACK for each decoded packet correctly and either a NAK or no ACK for each decoded packet with errors. Station A can use ACK / NAK feedback to relay decoded packets with errors. ACK / NAK feedback is not explicit feedback information regarding the recommendation of the speed (s) to be used for data transmission or how much to increase

or decrease the selected speed (s). For the sake of clarity, some parts of the following description assume that ACKs are returned for correctly decoded packets and that no ACKs (or that ACKs are absent) are returned for failed decoded packets.

FIG. 3 also shows an embodiment of an open loop speed control mechanism that includes an internal loop and an external loop. The internal loop estimates the SNRs observed at the receiving station B based on the pilot signals and / or the data received from the station B and the ASYM parameter, and selects a suitable speed for each data flow sent to the station B. The external loop adjusts the operation of the internal loop to achieve the desired level of performance, which can be quantified by a target packet error rate, for example, a PER of 1%.

For the internal loop, station B can transmit a pilot signal through link (B, A) to station A. In station B, the pilot signal is processed by a TX 382 space processor and sent from a transmitter unit

384. At station A, the pilot signal is received by a receiver unit 330 and is provided to a channel / processor estimator 332. Channel estimator 332 determines the SNRs received for the link (B, A) based on the signal pilot received and obtains SNR estimates for the link (B, A) based on the SNRs received. Speed selector 334 obtains SNR estimates for the link (A, B) based on SNR estimates for the link (B, A) and the ASYM parameter and also selects the speeds for data transmission on the link ( A, B).

5

10

fifteen

twenty

25

30

35

40

For the external loop, station B can send an ACK feedback to station A for packages correctly received from station A and / or a NAK feedback for packages received incorrectly or packages that are expected but not received. At station B, an ACK 380 generator generates ACK / NAK based on the packet status provided by the RX 368 data processor. Station B further processes and transmits the ACK / NAK to station A. At station A, an ACK / NAK 336 detector detects the transmitted ACK / NAKs and provides the detected ACK / NAKs to a speed adjustment unit 338. Unit 338 adjusts the ASYM parameter, the speeds and / or one or more speed adjustment parameters based on the ACK / NAKs detected and provides the settings to the speed selector 334. The speed setting parameters may include, for example, a SNR reduction factor, a setting for the ASYM parameter, an indication to reduce the number of data streams, an indication to reduce data stream rates, etc., as described below. Speed selector 334 uses the settings together with the SNR estimates and the ASYM parameter to select M speeds for the M data streams sent through the link (A, B).

Each station can maintain an internal loop and an external loop for the transmission of data in its direct link to another station. The internal loop can be activated by SNR estimates, and the external loop can be activated by ACK / NAK and / or other information. Each station can obtain and use SNR estimates to decide what speeds to use for data transmission to the other station. The accuracy of SNR estimates for the transmission link depends on the accuracy of the ASYM parameter. However, even in the absence of a good estimate for the ASYM parameter, the external loop can activate the speed selection by the transmitting station in a timely manner so that suitable speeds are used for data transmission.

The ability to accurately estimate SNRs for a link based on a transmission received through the other link allows you to select speeds that are closer to the higher speeds supported by the link. This allows each data stream to be sent with a smaller SNR margin, which is the difference between the SNR observed at the receiving station for that data stream and the SNR required for the speed selected for the data stream. Smaller SNR margins for data flows result in better utilization of link capacity. An accurate SNR estimate also avoids excess packet errors. In addition, the ability to use an SNR measurement for the reverse link to transmit data on the direct link allows rapid channel tracking. There is no need to transmit on the link (A, B), then wait for the speed feedback on the link (B, A) or finally transmit data at the appropriate speed on the link (A, B). In contrast, the SNR measurement for the link (B, A) can be used to transmit data on the link (A, B) and the round trip time can be quite short.

An open-loop speed control can be implemented in several ways and can be used for various transmission schemes. Example designs for the internal loop and for the external loop are described below.

1. Transmission modes

A MIMO system can support multiple transmission modes for better performance and greater flexibility. Table 2 lists some transmission modes as an example.

Table 2

Transmission Mode

Description

oriented mode

M data streams are transmitted in M orthogonal space channels (or own modes) of a MIMO channel, where 1  M  S.

Transmission Mode

Description

non-oriented mode

M data streams are transmitted on M space channels, for example, from M transmission antennas.

spatial dispersion mode

M data streams with spatial dispersion are transmitted to achieve similar performance for all data streams.

Each mode of transmission has different capacities and requirements. The oriented mode (which is also called own orientation) transmits data in the modes of a MIMO channel. The oriented mode normally provides better performance but requires spatial processing at both the transmitting and receiving stations. The non-oriented mode does not require any spatial processing at the transmitting station, for example, a data stream can be sent from each transmitting antenna. Normally, the performance of the non-oriented mode is not as good as that of the oriented mode. The spatial dispersion mode transmits M data flows with different orientation matrices so that these flows observe a set of effective channels and achieve similar performance. A suitable transmission mode can be selected for use depending on the availability of the channel status information (CSI), the capabilities of the transmitting and receiving stations, and so on.

For the oriented mode, the data is transmitted in S own modes, at most, of the MIMO channel formed by T transmit antennas and R receive antennas. The MIMO channel can be characterized by an R HT channel response matrix H, which can be diagonalized to obtain the S own modes of H. This diagonalization can be achieved by carrying out either a singular value decomposition of H or a decomposition in eigenvalues of a correlation matrix of H, which is R = HHH, where HH denotes the transposed conjugate of H. Decomposition into eigenvalues of R can be expressed as:

image4

where

E is a unitary matrix TT of own vectors of R; Y

 is a diagonal matrix TT of eigenvalues of R.

A unitary matrix E is characterized by the property EHE = I, where I is the identity matrix. The columns of a unit matrix are orthogonal to each other and each column has the unit power. The own vectors in E can be used (1) for spatial processing by the transmitting station to transmit data in the S own modes of H and (2) for spatial processing of the receiver by the receiving station to recover the data transmitted in the S own modes. The diagonal matrix  contains real non-negative values along the diagonal and zeros in the other positions. These diagonal entries are referred to as the proper values of H and represent the power gains for the S own modes of H.

For a MIMO-OFDM system, a channel response matrix H (k) can be obtained for each subband k and decomposed to obtain the eigenvectors E (k) and the eigenvalues  (k) for that subband. The S own values for each subband k can be ordered from the largest to the smallest, and the S own modes of H (k) will therefore be ordered from the highest SNR to the lowest SNR. A proprietary mode m of broadband, for m = 1, ..., S, can be formed with its own mode m for each of the K subbands. The main own broadband mode (with m = 1) is associated with the largest own value for each of the K subbands, the second own broadband mode (with m = 2) is associated with the second own plus value Great for each of the K subbands and so on.

Table 3 summarizes the spatial processing at the transmitting station and at the receiving station for the three transmission modes provided in Table 2. The same processing is performed for each subband, and the subband index k is not shown in the Table 3.

Table 3

oriented mode

non-oriented mode Spatial dispersion

Transmitter

xes = E · s xus = s xss = V · s

Effective channel

Hes = H · E Hus = H Hss = H · V

Symbols Received

rx = H · xx + n = Hx · s + n

Total CSI Receiver /

Month = Λ – 1 · EH · HH

CCMI

image4

MMSE receiver

image4

In Table 3, the subscript "is" denotes the oriented mode (or own orientation), "us" denotes the non-oriented mode, "ss" image4

denotes the spatial dispersion mode and "x" can be "is", "us", or "ss". is the variance of the AWGN noise in the channel

5 MIMO. s is a vector T1 with M data symbols to be sent in a subband in a symbol period and T - M zeros. V is a TT orientation matrix used for spatial dispersion. xx is a vector T1 with T transmission symbols to be sent from the T transmission antennas in a subband in a symbol period for the transmission mode x. Hx is an effective R respuestaT channel response matrix for the transmission mode x. M is a TR spatial filter matrix that can be obtained using a total CSI technique, a matrix inversion technique of image4

10 channel correlation (CCMI) or a minimum mean square error (MMSE) technique. It is a vector T1 of detected data symbols and is an estimate of s. MMSE and CCMI techniques can be used for all three modes of transmission. The MMSE and CCMI spatial filter matrices can be obtained in the same way for all three modes of transmission, although with different effective channel response matrices Hes, Hus and Hss.

Successive cancellation of interference can also be used to recover the M data flows in M stages.

15 Each stage  detects a data stream using one of the receiver spatial processing techniques shown in Table 3, decodes the detected data symbol stream, estimates the interference caused by the decoded data stream in the non-data streams still recovered and cancels the estimated interference of the received symbol flows to obtain modified symbol flows for the next step  +1. Then, the next step  +1 will detect, decode and cancel another data stream.

20 The following describes the speed selection and open loop speed control for data transmission from station A to station B in a MIMO-OFDM TDD system.

2. Internal loop

The internal loop selects the speeds for the M data streams sent by station A to station B. The components of the internal loop include SNR estimation, power allocation and speed selection.

25 A. SNR estimation

The SNR received for each subband of each transmission channel on the link (B, A) can be determined based on a pilot signal received through the link (B, A) and the receiver spatial processing technique used by the station A. The pilot signal may be a "clean" pilot signal sent without any spatial processing, a "oriented" pilot signal sent in the proprietary modes with E, or a "scattered" pilot signal sent on multiple spatial channels

30 with V.

For the total CSI technique for the oriented mode, the SNR received for each subband k of each own mode m of broadband,  is, m (k), can be expressed as:

image4

where Pm (k) is the transmission power used for subband k in its own way m broadband; and m is the m-th diagonal element of  (k) for subband k.

For the MMSE technique for the three modes of transmission, the SNR received for each subband k of each broadband space channel m, mmse, m (k), can be expressed as:

image4

where qm (k) is the m-th diagonal element of

image4 for the subband k.

For the CCMI technique for the three modes of transmission, the SNR received for each subband k of each broadband space channel m, ccmi, m (k), can be expressed as:

image4

where the subband rods k.

In equations (9), (10) and (11), linear units The quantities es, m (k), mmse, m (k) and ccmi, m (k) are the SNRs after the spatial processing of receiver, are in units of decibels (dB) and are also referred to as the SNRs received.

For the spatial dispersion mode, the S spatial channels get similar received SNRs due to the dispersion spatial with different V matrices. Therefore, the same speed can be used for all data flows sent on these space channels. With spatial dispersion, an average SNR for all spatial channels It can be expressed as:

image5

image4

The calculation of the SNR mean can be done in linear units, as shown in equations (12) and (13),

or in units of dB.

The SNR received from each data stream can also be determined based on the data symbols detected for that stream. For data-based SNR estimation, the receiving station performs spatial processing on the data symbols received with the spatial filter matrix for each subband to obtain the detected data symbols. Each data symbol detected is normally a multi-bit value (or a flexible decision) that is an estimate of a transmitted data symbol. The receiving station can convert the detected data symbols into firm decisions, determine the difference between flexible decisions and firm decisions for each data symbol detected in each data flow and calculate an average error power for each data flow such as the mean squares of magnitude of the difference between flexible decisions and firm decisions for data flow. Then, the receiving station can obtain a received SNR based on data for each data stream based on the average error power for that data stream.

The pilot signal is normally transmitted so that the SNRs received based on pilot signals can be obtained for all S space channels, even if the data is transmitted on less than S space channels. Received SNRs based on data can be obtained for the M data streams that are actually transmitted. Both received SNRs based on pilot signals and received SNRs based on data can be used to select speeds for data transmission. A discrepancy between the SNRs received based on pilot signals and those based on data may be indicative of deficiencies in the transmitting station and / or in the receiving station. In this case, the lower or the average of the two types of SNR received can be used to select the speed for each data stream.

B. Power Assignment

The SNRs received for each transmission channel (for example, each proprietary broadband mode or broadband spatial channel) can vary greatly across the total K subbands. Better performance can be achieved by using certain subbands for data transmission and / or by allocating the available transmission power in different ways.

In one embodiment, only valid subbands are used for data transmission and inappropriate subbands are not used for data transmission. For this embodiment, the SNR received m (k) for each subband can be determined based on a nominal transmission power for the subband, for example, with the total transmission power for the transmission channel distributed evenly across the total K subbands. A subband may be considered as (1) a valid subband if its received SNR exceeds a low SNR threshold, m (k)> umb_bajo, or as

(2) an inadequate subband, otherwise. The total transmission power for the transmission channel is distributed evenly across valid subbands.

In another embodiment, only valid subbands are used for data transmission, inappropriate subbands are not used for data transmission, and the total transmission power for the transmission channel is distributed unevenly across valid subbands. . For this embodiment, each subband may be considered as a valid subband or as an inadequate subband. Then, the total transmission power for the transmission channel is distributed evenly across the valid subbands. The SNR received for each valid subband is determined based on the transmission power assigned to that subband. If the received SNR exceeds a high SNR threshold, m (k)> umb_alto, where umb_alto> umb_bajo, then a sufficient transmission power is assigned to the subband to achieve the high SNR threshold, and the excess transmission power The total excess transmission power collected from all valid subbands with SNRs received greater than umb_high is then redistributed afterwards (for example, uniformly) through the other valid subbands with SNRs received less than umb_high. The process can be repeated until all excess power is exhausted. This embodiment collects transmission powers of inadequate subbands and the excess of transmission powers of very valid subbands and redistributes the transmission powers collected through the valid subbands.

In another additional embodiment, the channel inversion is performed so that all subbands get similar SNRs received. For this embodiment, the amount of transmission power allocated to each subband is inversely related to the SNR received for that subband. Subbands with lower received SNRs are assigned more transmission powers, and subbands with higher received SNRs are assigned less transmission powers. This embodiment may improve performance for certain coding schemes that prefer less variation in the SNRs received through a packet.

In another additional embodiment, only valid subbands are used for data transmission, inappropriate subbands are not used for data transmission, and channel inversion is performed through valid subbands. For this embodiment, valid subbands can first be identified based on the SNR threshold umb_bajo. Then, the total transmission power for the transmission channel is distributed unevenly across the valid subbands so that those subbands get similar received SNRs.

Subbands can also be selected for data transmission in other ways. The total transmission power for each transmission channel can also be distributed through the subbands selected in other ways.

For the embodiments described above, the SNR thresholds umb_bajo and umb_alto can be fixed values. Alternatively, these SNR thresholds may be a function of the speed used for the transmission channel and / or some other factors. For example, lower values can be used for SNR thresholds for lower speeds, and higher values can be used for SNR thresholds for higher speeds.

The power allocation described above can improve performance without requiring explicit feedback from the receiving station. The transmitting station can determine the received SNRs for the subbands of each transmission channel based on the pilot signal of the receiving station and can use those received SNRs to select subbands for data transmission and to distribute the total transmission power to the subbands selected. The receiving station may not need to know the selection of subbands or the distribution of transmission power by the transmitting station.

C. Speed selection

The speed for each data stream can be selected as follows. The SNR received for each subband k used for each data stream m can be calculated as described above and denoted as m (k). For a MIMO-OFDM system, the SNRs received for each data flow normally depend on the frequency and also depend on the mode of transmission and the spatial processing technique of the receiver used for data transmission. In the following description, all amounts of SNR are in units of dB.

The average SNR for each data stream m, media, m, can be calculated as:

image4

The variance of the SNRs for each data flow m, image4 It can be calculated as:

image4

An SNR variability recoil factor for each data stream m, os, m, which is used to represent the variability in the SNRs received through the K subbands can be calculated as:

image4

where Kos is a constant used to reduce the estimate of SNR from the average SNR to reflect the fact that many error correction codes do not handle a large variation of SNR well through a packet to be decoded. The Kos factor can be selected based on the error correction code and / or the code rate used for the data flow m. In general, the SNR variability regression factor can be calculated based on any function of the mean SNR and SNR variance, or

image6

It can be defined as:

where SNR estimation to account for the speed mismatches that may occur due to (1) noise in the link and in an SNR estimation process and (2) to a misalignment of the orientation vectors used by the transmitting stations and receivers for oriented mode. The factor  is normally a positive constant near

image4

to the unit (1,0). The order of diversity Kd for the oriented mode can be given as:

image4

The order of diversity Kd represents the amount of spatial diversity achieved for data transmission. The order of diversity increases as more reception antennas are used for a given number of data streams. A higher diversity order usually corresponds to a smaller variation in the SNRs received for each data flow.

The flow rate m of equation (17) represents the fact that different data flows may have different amounts of SNR variability for the oriented mode. The main own broadband mode has the highest average SNR and the lowest SNR variability over time and frequency, and a smaller MIMO reduction factor (with m = 1) can be used for this own broadband mode . The S – th own broadband mode has the lowest average SNR and more SNR variability over time and frequency, and a larger MIMO feedback factor (with m = S) can be used for this band-specific mode wide The flow rate m can also be omitted from equation (17).

A MIMO recoil factor for the non-oriented mode and for the spatial dispersion mode can be defined as:

image4

where Kd (m) is the order of diversity observed by the data flow m, which is:

image4

With a successive cancellation of interference, after estimating and canceling the interference due to the data flow m, the order of diversity for each subsequent data flow increases as if the data flow m had not been transmitted. For example, if R = 4 and M = 3, then the data flow m = 1 observes an order of diversity of two, the data flow m = 2 observes an order of diversity of three and the data flow m = 3 Look at an order of diversity of four.

The SNR estimate for each data flow can be calculated as:

image4

where

 external, m is an external loop reduction factor for the data flow m;

m, m is a MIMO recoil factor for the data flow m; Y

est, m is the estimate of SNR for the data flow m.

The minimum MIMO recoil factor, m can be determined as shown in equation (17) or (19) and can be used to represent various characteristics associated with MIMO transmission. In general, the various factors described above (for example, Kos, Kd and ) can be selected to maximize performance and minimize the packet error rate and can be determined based on a calculation simulation, empirical measurements, etc. The SNR estimate can also be obtained based on other factors. For example, an L factor can be used to represent variations in packet length. A larger package generally requires a higher SNR to get a given PER.

For an open loop speed control, the transmitting station A can obtain the SNR estimates for the receive link (B, A) as described above. Then, station A can obtain the SNR estimates for the transmission link (A, B) based on the SNR estimates for the link (B, A) and the ASYM parameter, as shown in equation ( 6) or (7). Station A can use SNR estimates for link (A, B) to select the speeds to be used for the data streams sent on link (A, B) to station B.

5 In one embodiment, the speed for each data stream is independently selected based on the SNR estimate for that data stream. The system can support a set of speeds. Table 4 lists an example set of 14 speeds supported by the system. Each speed is associated with a specific spectral efficiency, a specific code rate, a specific modulation scheme and a specific minimum SNR required to achieve a PER of 1% for an AWGN channel without signal fading. Efficiency

10 spectral refers to a data transfer rate normalized by the system bandwidth and is given in units of bits per second per hertz (bps / Hz). The code rate and modulation scheme for each speed in Table 4 are specific to the system design as an example. For each non-zero rate, the required SNR is obtained based on the specific system design (for example, the code rate, the interlacing scheme and the modulation scheme used for that rate) and for an AWGN channel. The required SNR can

15 Obtained through a calculation simulation, a computer simulation, empirical measurements, etc.

Table 4

Speed index

Spectral Efficiency (bps / Hz) Code speed Modulation scheme SNR required (dB)

0

0.0 - - -

one

0.25 ¼ Bpsk –1.8

2

0.5 ½ Bpsk 1.2

3

1.0 ½ QPSK 4.2

4

1.5 ¾ QPSK 6.8

5

2.0 ½ 16 QAM 10.1

6

2.5 5/8 16 QAM 11.7

7

3.0 ¾ 16 QAM 13.2

8

3.5 12/7 64 QAM 16.2

9

4.0 2/3 64 QAM 17.4

10

4,5 ¾ 64 QAM 18.8

eleven

5.0 5/6 64 QAM 20.0

12

6.0 ¾ 256 QAM 24.2

13

7.0 7/8 256 QAM 26.3

A query table can be used to store the set of supported speeds and the SNR required for each supported speed. The SNR estimate for each data flow can be provided to the query table and

20 compared to the SNR required for the supported speeds. Then, the query table provides a selected rate for each data stream, which is the rate supported with the highest performance and a required SNR that is less than or equal to the SNR estimate for that data stream.

In another embodiment, the speeds for the M data streams are selected with a margin sharing. The speed for each data stream is initially selected based on the SNR estimate for that data stream,

25 as described above. The SNR margin for each data flow is determined. The total SNR margin is then calculated and distributed among the data streams so that one or more speeds can be increased for one or more data streams.

In another additional embodiment, the speeds for the M data streams are selected jointly based on the SNR estimates for those data streams. The system can support a set of vector-quantified speeds in which only certain combinations of speeds are allowed. Each combination of speed indicates a particular number of data streams to be transmitted and the rate to be used for each data stream. A combination of speed with the highest overall performance can be selected based on SNR estimates for data flows.

The number of data streams to be transmitted (M) can also be selected based on SNR estimates. In one embodiment, the overall performance is calculated for each of the possible numbers of data streams, for example, for M = 1, 2, ..., S. For each value of M, the total transmission power is distributed (by for example, uniformly) through M data streams, the received SNRs are calculated for each data stream based on the assigned transmission power, the SNR estimate is obtained for each data stream and is used to select the speed for that data stream, and the overall performance is calculated as the sum of the data rates selected for the M data streams. The highest overall performance is determined among the S global returns calculated for the S different possible numbers of data streams, and the number of data streams that provides this highest overall performance is selected as M.

For the spatial dispersion mode, the optimal number of data streams is rarely equal to the number of spatial channels. The simultaneous transmission of S data flows normally results in excessive crosstalk. Therefore, the overall performance that can be achieved with S data streams is normally lower than the overall performance that can be achieved with S – 1 or less data streams. For the spatial dispersion mode, M may be limited as M ≤ S – 1.

The transmission mode to be used for data transmission can be selected depending on the age of the orientation vectors. If the orientation vectors are recent enough, then the data streams transmitted with these targeting vectors will be suitable for the wireless channel and the receiving station will receive orthogonal data streams. As the mismatches between the orientation vectors and the wireless channel increase, the gain of the own orientation decreases due to greater crosstalk between the data streams. Under these conditions, it may be more appropriate to use the spatial dispersion mode for data transmission. The transmission mode can also be selected based on other factors such as, for example, the capabilities of the transmitting and receiving stations, the coherence time of the MIMO channel, the feedback delay, the type of data traffic to which it is going to give service, etc.

3. External loop

In one embodiment, the ASYM parameter is fixed and the external loop adjusts the external loop recoil factor for each data stream to achieve the target PER for that data stream. The external loop recoil factor can be maintained independently for each data stream if the packets for each data stream are coded separately and the feedback is per packet or per burst. The external loop backward factor for each data stream m can be updated as follows:

image4

where m is an increase value of the external loop reduction factor for a deleted packet; Pm is the fraction of m to be subtracted for a valid package; Y external, m (n) is the external loop backward factor for a packet n of the data stream m. C is a constant value that is usually positive but can also be negative. The increment value m determines the convergence rate for the external loop. Loopback factor

external can be initialized at the beginning of the data transmission at a predetermined value (for example, to zero or some other value) and, therefore, can be updated depending on the status of the packets received depending on the ACK / NAK.

The factor Pm determines the packet error rate for the data flow m and can be expressed as: where PERm is the target PER for the data flow m. For example, if PERm = 0.01 for a PER of 1%, then Pm = 0.0101 = 1/99.

image4

For coded data streams, a single packet can be encoded, divided and sent through those data streams. In this case, it can be difficult to determine which of the data flows caused a packet error. A margin of SNR can be maintained for each data flow. The SNR margin for the data stream m is calculated as the difference between the SNR required for the selected rate and the SNR estimate for that data stream. For each packet deleted, it can be assumed that the data flow with the smallest SNR margin has caused the packet error, and the external loop backward factor for this data flow can be increased. For each valid packet, external loopback factors for all data flows can be decreased. If all data flows have practically identical SNR margins, which is usually the case for spatial dispersion mode, then a single external loopback factor can be maintained for all data flows.

In another embodiment, the external loop adjusts the ASYM parameter based on the ACKs to achieve the target PER. The ASYM parameter can be updated as follows:

image4

where ASYM is an increment value for the ASYM parameter for a deleted package.

The above description is for a specific embodiment of the internal and external loops. For this embodiment, the internal loop selects the transmission mode, the number of data streams and the speed for each data stream based on the received SNRs and other parameters. The external loop adjusts one or more of the parameters depending on the ACK feedback. Speed selection and speed control can also be carried out in other ways. For example, the external loop can control the internal loop to reduce or increase the number of data streams, adjust one or more speeds for one or more data streams, etc.

In another embodiment of the internal and external loops, the transmitting antenna A transmits a single data stream at a rate that can be selected based on the ASYM parameter, as described above. Station A determines the SNR margin for the data stream and selects the next highest rate for the data stream if the SNR margin is positive and the most recent packet transmission (s) is / are satisfactory (s). Station A allows an additional flow once the highest speed for current data flows has been reached. When the additional data flow is enabled, the overall performance is limited to a value X times (for example, X  1.3) greater than the previous global performance and is distributed (for example, uniformly) between the flows of data enabled Station A can select the next highest speed for one or both data streams if the SNR margin remains positive and packet transmissions remain satisfactory. Additional data streams can be enabled in a similar way. If the situation of a deleted packet occurs (for example, an ACK is not received), then station A can (1) reduce the speed for one or more data streams or select the last known combination of valid speeds and / or (2) reduce the number of data flows. For example, station A can reduce the speed if the received SNR is relatively static and can reduce the number of data streams if the received SNR has changed abruptly. Station A can also reduce the overall speed by some percentage (for example, 50%) if the reduction in speed and / or flows still results in packet errors.

Other designs for internal and external loops can also be implemented, and this is within the scope of the invention.

A system typically includes multiple access points and multiple user terminals. Each access point can periodically transmit a pilot signal (or a beacon) and other signaling. A user terminal can search for pilot signals from the access points and can access and register with each access point whose pilot signal has been received with sufficient intensity. Relevant information may be exchanged during registration to allow both the user terminal and the access point to determine the ASYM parameters for their links.

For uplink data transmission from the user terminal to the access point, the user terminal can obtain SNR estimates for the downlink based on the pilot signal periodically transmitted by the access point and / or the last transmission downlink to the user terminal. The user terminal can then obtain SNR estimates for the uplink based on the downlink SNR estimates and the ASYM parameter and can select one or more speeds for uplink transmission based on the SNR estimates of uplink

For the transmission of downlink data from the access point to the user terminal, the access point may not receive any transmission from the user terminal and may not be able to obtain uplink SNR estimates. In this case, the access point can start with a single data flow and can use a predetermined speed. In one embodiment, this speed may be an average speed (for example, 24 Mbps for IEEE 802.11a), the last known valid speed for the user terminal, the highest speed or some other speed. If control frames are exchanged before data transmission, then the access point may initially transmit a control frame at the predetermined speed and / or use the speeds required for the transmission of control frames (for example, one of the basic speeds of 6 Mbps, 12 Mbps or 24 Mbps for IEEE 802.11a) and can progressively reduce the speed until the control frame is correctly received. Then, the speed for data transmission can be selected based on the speed for the control frame and / or information obtained from a control response frame.

The speed control techniques described in this document can be implemented in several ways. For example, these techniques can be implemented in hardware, software or in a combination thereof. For a hardware implementation, the processing units used to perform the speed control in a transmitting station (for example, the channel estimator 322, the speed selector 334, the speed adjustment unit 338, etc.) can be implemented in one or more application specific integrated circuits (ASICs), digital signal processors (DSP), digital signal processing devices (DSPD), programmable logic devices (PLD), field programmable door arrays (FPGA), processors , controllers, microcontrollers, microprocessors, other electronic units designed to perform the functions described in this document, or a combination thereof, each of which may be coupled as appropriate in one

or more devices The processing units used to support speed control in a receiving station can also be implemented in one or more ASICs, DSPs, processors, etc.

For a software implementation, speed control techniques can be implemented with modules (for example, procedures, functions, etc.) that perform the functions described in this document. The software codes can be stored in a memory unit (for example, memory unit 342 of FIG. 3) and executed by a processor (for example, controller 340 of FIG. 3). The memory unit can be implemented within the processor or externally to the processor, in which case it can be communicatively coupled to the processor through various means, as is known in the art.

The headings are included in this document by way of reference and to help locate certain sections. These headings are not intended to limit the scope of the concepts described therein, and these concepts may be applied to other sections throughout the entire specification.

The above description of the disclosed embodiments is provided for any person skilled in the art to carry out or use the present invention. Various modifications of these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein can be applied to other embodiments without departing from the scope of the invention. Therefore, the present invention is not intended to be limited to the embodiments shown in this document but is granted the broadest scope compatible with the novel principles and features disclosed in this document.

Claims (36)

1. A procedure for making a speed selection, comprising: estimate (114) the channel quality of a first communications link based on a transmission received through the first communications link; estimate (116) the channel quality of a second communications link based on the estimated channel quality of the first communications link and an asymmetric parameter indicative of a difference in the channel qualities of the first and second communications link, where the difference in channel qualities is the difference between an SNR for the first communications link and an SNR for the second communications link; Y select (118) at least one speed for a data transmission through the second communication link based on the estimated channel quality of the second communication link. 2. The method according to claim 1, wherein: the estimate (114) of the channel quality of the first communications link comprises obtaining (216) an SNR estimate for a first transmission channel in the first communications link based on the transmission received through the first communications link, the estimation (116) of the channel quality of the second communication link comprises obtaining (218) an estimate of SNR for a second transmission channel in the second communication link based on the estimation of SNR for the first transmission channel and of the asymmetric parameter, and the selection (116) of the at least one speed comprises selecting (220) a speed for the second transmission channel based on the estimation of SNR for the second transmission channel.

3.

The method according to claim 2, further comprising:

determine at least one received SNR for at least one frequency subband of the first transmission channel and obtain the SNR estimate for the first transmission channel based on the at least one received SNR.

Four.

The method according to claim 2, wherein obtaining the SNR estimate for the first transmission channel is further based on an order of diversity for the first transmission channel.

5.

The method according to claim 2, wherein obtaining the SNR estimate for the first transmission channel is further based on a factor representing the variability in the at least one received SNR.

6.

The method according to claim 2, further comprising:

determine at least one SNR received based on a receiver spatial processing technique used by a receiving station for data transmission.

7.

The method according to claim 1, further comprising:

determine the asymmetric parameter based on the capabilities of a transmitting station and the capabilities of a receiving station for data transmission.

8.

The method according to claim 7, wherein the capacities of both the transmitting station and the receiving station comprise a maximum transmission power, a receiver noise factor and a plurality of reception antennas.

9.

The method according to claim 1, further comprising:

determine the asymmetric parameter based on an SNR received for the first communications link and an SNR received for the second communications link.

10.

The method according to claim 1, further comprising: adjusting (120) the asymmetric parameter according to the performance of the data transmission.

eleven.

The method according to claim 1, further comprising:

Receive feedback for packets sent for data transmission and adjust the asymmetric parameter based on the feedback received.

12.

The method according to claim 1, wherein selecting the at least one speed is further based on a recoil factor.

13.

The method according to claim 12, further comprising:

Receive feedback for packets sent for data transmission and adjust the feedback factor based on the feedback received.

14.

The method according to claim 1, wherein adjusting the at least one speed, modifying the number of data streams or both actions are based on the performance of the data transmission.

fifteen.

The method according to claim 1, further comprising: adjusting a recoil factor based on the performance of the data transmission, and wherein the at least one speed is further selected based on the recoil factor.

16.

The method according to claim 1,

wherein estimating the channel quality of the first communications link comprises obtaining a first plurality of SNR estimates for a first plurality of transmission channels in the first communications link based on the transmission received through the first communications link; wherein estimating the channel quality of the second communications link comprises obtaining a second plurality of SNR estimates for a second plurality of transmission channels in the second communications link based on the first plurality of SNR estimates and the asymmetric parameter ; Y wherein selecting at least one speed for data transmission comprises selecting a plurality of speeds for the second plurality of transmission channels based on the second plurality of SNR estimates. 17. The method according to claim 16, wherein obtaining the first plurality of SNR estimates comprises determine SNRs received for a plurality of frequency subbands of each of the first plurality of transmission channels; Y Obtain an SNR estimate for each of the first plurality of transmission channels based on the SNRs received.

18.

The method according to claim 16, further comprising:

spatially process a plurality of data streams for an oriented mode, an unoriented mode or a spatial dispersion mode.

19.

An apparatus, comprising:

means for estimating the channel quality of a first communications link based on a transmission received through the first communications link; means for estimating the channel quality of a second communications link based on the estimated channel quality of the first communications link and an asymmetric parameter indicative of a difference in the channel qualities of the first and second communications link, where the difference in channel qualities is the difference between an SNR for the first communications link and an SNR for the second communications link; Y means for selecting at least one speed for a data transmission through the second communication link based on the estimated channel quality of the second communication link. 20. The apparatus according to claim 19, wherein: The means for estimating the channel quality of the first communications link comprise means for obtaining an SNR ratio estimate for a first transmission channel in the first communications link based on the transmission received through the first communications link; The means for estimating the channel quality of the second communications link comprise means for obtaining an SNR estimate for a second transmission channel on the second communications link based on the SNR estimate for the first transmission channel and the asymmetric parameter. , Y The means for selecting the at least one speed comprise means for selecting a speed for the second transmission channel based on the estimation of SNR for the second transmission channel.

twenty-one.

The apparatus according to claim 20, further comprising:

means for determining at least one received SNR for at least one frequency subband of the first transmission channel and for obtaining the SNR estimate for the first transmission channel based on the at least one received SNR.

22

The apparatus according to claim 20, wherein the means for obtaining the SNR estimate for the first transmission channel obtains the SNR estimation based on an order of diversity for the first transmission channel.

2. 3.

The apparatus according to claim 20, wherein the means for obtaining the SNR estimate for the first transmission channel obtains the SNR estimate based on a factor that represents the variability in the at least one received SNR.

24.

The apparatus according to claim 20, further comprising:

means for determining at least one SNR received based on a receiver spatial processing technique used by a receiving station for data transmission.

25.

The apparatus according to claim 19, further comprising:

means for determining the asymmetric parameter based on the capabilities of a transmitting station and the capabilities of a receiving station for data transmission.

26.

The apparatus according to claim 25, wherein the capacities of both the transmitting station and the receiving station comprise a maximum transmission power, a receiver noise factor and a plurality of reception antennas.

27.

The apparatus according to claim 19, further comprising:

means for determining the asymmetric parameter based on an SNR received for the first communications link and an SNR received for the second communications link.

28.

The apparatus according to claim 19, further comprising: means for adjusting the asymmetric parameter according to the performance of the data transmission.

29.

The apparatus according to claim 19, further comprising:

means for receiving feedback for packets sent for data transmission and for adjusting the asymmetric parameter based on the feedback received.

30

The apparatus according to claim 19, wherein selecting the at least one speed is further based on a recoil factor.

31.

The apparatus according to claim 30, further comprising:

means for receiving feedback for packets sent for data transmission and for adjusting the recoil factor based on the feedback received.

32

The apparatus according to claim 19, wherein adjusting the at least one speed, modifying the number of data streams or both actions are based on the performance of the data transmission.

33.

The apparatus according to claim 19, further comprising:

means for adjusting a recoil factor based on the performance of the data transmission, and in which the at least one speed is further selected based on the recoil factor.

3. 4.

The apparatus according to claim 19,

in which the means for estimating the channel quality of the first communications link comprise means 5 to obtain a first plurality of SNR estimates for a first plurality of transmission channels in the first communications link depending on the transmission received through the first communications link; wherein the means for estimating the channel quality of the second communications link comprises means for obtaining a second plurality of SNR estimates for a second plurality of channels 10 on the second communication link based on the first plurality of SNR estimates and the asymmetric parameter; Y wherein the apparatus further comprises means for selecting a plurality of speeds for the second plurality of transmission channels based on the second plurality of SNR estimates. 35. The apparatus according to claim 34, wherein the means for obtaining the first plurality of SNR estimates comprises means for determining SNRs received for a plurality of frequency subbands of each of the first plurality of transmission channels; Y means for obtaining an SNR estimate for each of the first plurality of transmission channels based on the SNRs received. The apparatus according to claim 34, further comprising: means for encoding and modulating a plurality of data flows based on the plurality of speeds. 37. The apparatus according to claim 34, further comprising: means for spatially processing a plurality of data streams for an oriented mode, an unoriented mode or a spatial dispersion mode.