HK1105489A - Efficient computation of spatial filter matrices for steering transmit diversity in a mimo communication system - Google Patents

Description

Efficient computation of spatial filter matrices for steering transmit diversity in a MIMO communication system

Technical Field

The present disclosure relates generally to communication, and more specifically to spatial processing for data transmission in a multiple-input multiple-output communication system.

Technical Field

MIMO systems employ multiple (N) at the transmitting entity_T) Transmitting antennas and employing multiple (N) at a receiving entity_R) And receiving the antenna for data transmission. Can be substituted by N_TA transmitting antenna and N_RMIMO channel formed by multiple receiving antennas is decomposed into N_SA spatial channel of which N_S≤min{N_T，N_R}. This N_SThe spatial channels may be used to transmit data in parallel to improve throughput and/or redundantly to improve reliability.

Each spatial channel may be subject to various harsh channel conditions such as fading, multipath, and interference effects. N is a radical of_SThe spatial channels may also experience different channel conditions and may achieve different signal-to-noise-and-interference ratios (SNRs). The SNR of each spatial channel determines its transmission capacity, which is typically quantified by a particular data rate at which reliable transmission occurs over that spatial channel. For time-varying wireless channels, the channel conditions change over time, and the SNR of each spatial channel also changes over time.

To improve performance, MIMO systems may utilize some form of feedback, such that a receiving entity estimates spatial channels and provides feedback information indicative of the channel conditions or transmission capacity of each spatial channel. The transmitting entity may then adjust the data transmission on each spatial channel based on the feedback information. However, feedback information may not be available for various reasons. For example, the system may not support feedback transmissions from the receiving entity, or the wireless channel may change faster than the receiving entity can estimate the wireless channel and/or send back feedback information. In any case, if the transmitting entity does not know the channel conditions, it needs to transmit data at a low rate so that the receiving entity can reliably decode the data transmission even in the worst channel conditions. The performance of the system will be limited by the expected worst channel conditions, which is highly undesirable.

To improve performance (e.g., when feedback information is not available), the transmitting entity may perform spatial processing such that the data transmission does not observe the worst channel conditions for an extended period of time, as described below. The higher data rate can then be used for data transmission. However, this spatial processing introduces additional complexity to the transmitting and receiving entities.

Therefore, there is a need in the art for techniques to efficiently perform spatial processing to improve performance in MIMO systems.

Disclosure of Invention

Techniques for efficiently computing a spatial filter matrix for spatial processing by a receiving entity are described below. The transmitting entity may transmit data via the MIMO channel using full channel state information ("full CSI") or "partial CSI" transmission, as described below. The transmitting entity may also utilize pilot transmit diversity (STD) to improve performance. With STD, the transmitting entity performs spatial processing using different steering matrices so that the data transmission observes all valid channels and does not block on a "bad" channel realization for an extended period of time. The receiving entity performs complementary receiver spatial processing for full-CSI or partial-CSI transmission and for pilot transmit diversity. The spatial filter matrix for receiver spatial processing can be efficiently computed if the MIMO channel is relatively static or does not change abruptly.

The channel response matrices for a MIMO channel are highly correlated over a series of transmission spans (e.g., a series of symbol periods or frequency subbands) if the MIMO channel is relatively static. In this case, an initial spatial filter matrix may be derived based on the channel response matrix and the selected receiver processing technique, as described below. A spatial filter matrix for each transmission span in the static range may then be calculated based on the initial spatial filter matrix and the steering matrix for that transmission span.

If the MIMO channel is not static but does not change abruptly, the channel response matrices for different transmission spans are partially correlated. In this case, a spatial filter matrix may be derived for a given transmission span lM _x(l) The matrix can be used to derive a matrix for another transmission span mAn initial spatial filter matrix. Then, a spatial filter matrix for the transmission span m is calculated based on the initial spatial filter matrix, e.g. using an iterative procedureM _x(m) of the reaction mixture. The same process may be repeated over a series of related transmission spans so that each newly derived spatial filter matrix may be used to compute another spatial filter matrix for another transmission span.

The steering matrix may be defined so that the computation of the spatial filter matrix may be simplified. Various aspects and embodiments of the invention are described in further detail below.

Drawings

Fig. 1 shows a transmitting entity and a receiving entity in a MIMO system;

FIG. 2 shows a model for data transmission using steered transmit diversity;

FIGS. 3A and 3B illustrate data transmission in a single carrier MIMO system and a multi-carrier MIMO system, respectively;

FIGS. 4 and 5 illustrate the process of computing spatial filter matrices for fully-correlated and partially-correlated channel response matrices, respectively;

FIG. 6 is a block diagram illustrating an access point and a user terminal; and

fig. 7 is a block diagram illustrating a processor for computing a spatial filter matrix.

Detailed Description

The word "exemplary" is used herein to mean "serving as an example, instance, or illustration. Any embodiment described herein as "exemplary" is not necessarily to be construed as preferred or advantageous over other embodiments.

FIG. 1 shows a MIMO system 100A simple block diagram of the transmitting entity 110 and the receiving entity 150. In transmitting entity 110, a Transmit (TX) spatial processor 120 applies vectors to data symbolss(m) to generate transmit symbols (in terms of m)x(m) represents). As used herein, a "data symbol" is a modulation symbol for data, a "pilot symbol" is a modulation symbol for pilot information (i.e., data known a priori by both the transmitting entity and the receiving entity), a "transmit symbol" is a symbol transmitted from a transmit antenna, a "receive symbol" is a symbol obtained from a receive antenna, and a modulation symbol is a complex value for a point in a signal constellation for a certain modulation scheme (e.g., M-PSK, M-QAM, etc.). Based on a steering matrixv(m) and possibly other matrices. A transmitter unit (TMTR)122 further conditions the transmitted symbols to generate N_TModulation symbols from N_TThe transmit antennas 124 transmit and via a MIMO channel.

In the receiving entity 150, the transmitted modulated signal consists of N_RA receiving antenna 152 receives, and N is_REach received signal is conditioned by a receiver unit (RCVR)154 to obtain received symbols (using vectors)r(m) represents). The spatial filter matrix is then used (by) a Receive (RX) spatial processor 160M _x(m) performs receiver spatial processing (or spatial matched filtering) on the received symbols to obtain "detected" data symbols (in vectors)Representation). The detected data symbols are estimates of the data symbols sent by transmitting entity 110. Spatial processing at the transmitting and receiving entities is described below.

The spatial filter matrix computation techniques described herein may be used for single carrier MIMO systems as well as multi-carrier MIMO systems. The multiple carriers may be obtained using Orthogonal Frequency Division Multiplexing (OFDM), Discrete Multitone (DMT), other multi-carrier modulation techniques, or other architectures. OFDM effectively partitions the overall system bandwidth into multiple (N)_F) Orthogonal sub-bands, which are also providedReferred to as tones, subcarriers, bins, and frequency channels. With OFDM, each subband is associated with a respective subcarrier that may be modulated with data.

In MIMO system 100, N at transmitting entity 110_TN at one transmit antenna and receiving entity 150_RThe MIMO channel formed by the receiving antennas can use N_R×N_TChannel response matrixH(m), which can be expressed as:

formula (1)

Wherein the term h_i，j(m) denotes the coupling or complex channel gain between the transmit antenna j and the receive antenna i for a transmission span m, i 1.. N_R，j，＝1...N_T. The transmission span may cover the time and/or frequency dimensions. For example, for a single carrier MIMO system, a transmission span may correspond to one symbol period, which is a time interval in which one data symbol is transmitted. For a multi-carrier MIMO system, a transmission span may correspond to one subband in one symbol period. A transmission span may also cover multiple symbol periods and/or multiple subbands. For simplicity, it is assumed that the MIMO channel is full rank, with N_S=N_T≤N_R。

MIMO systems may support data transmission in one or more operating modes (e.g., "calibration" and "non-calibration" modes). The calibration mode may employ full CSI transmission such that data is transmitted on orthogonal spatial channels (or "eigenmodes") of the MIMO channel. The non-calibration mode may employ partial CSI transmission such that data is transmitted on spatial channels of the MIMO channel, e.g., from a single transmit antenna.

MIMO systems may also employ Steered Transmit Diversity (STD) to improve performance. With STD, the transmitting entity performs spatial processing using the steering matrix so that the data transmission observes all valid channels and does not block on a single bad channel realization for an extended period of time. Thus, performance is not limited by the worst channel conditions.

1. Calibration mode-full CSI transmission

For full CSI transmission, can be pairedH(m) performing eigenvalue decomposition on the correlation matrix to obtainHN of (m)_SThe eigenmodes are as follows:

R(m)＝ H ^H(m)· H(m)＝ E(m)· Λ(m)· E ^H(m), formula (2)

Wherein the content of the first and second substances,R(m) isH(m) one N_T×N_TA correlation matrix;

E(m) is a number N_T×N_TUnitary matrix of which each column isR(m) eigenvectors;

Λ(m) isR(m) an eigenvalue of_T×N_TA diagonal matrix; and

“^H"denotes conjugate transpose.

Unitary matrixUPassing propertyU ^H· U＝ IIs characterized in thatIIs an identity matrix. The columns of the unitary matrix are orthogonal to each other and each column has a unit power. Matrix arrayE(m) can be spatially processed by the transmitting entity, therebyHN of (m)_SData is transmitted on one eigenmode. Eigenmodes can be considered as orthogonal spatial channels obtained by decomposition.ΛThe diagonal term of (m) isRThe eigenvalues of (m) represent N_SPower gain of the eigenmodes.Singular value decomposition may also be performed to obtain a matrix of left and right eigenvectors that may be used for full CSI transmission.

The transmitting entity performs spatial processing on the full CSI transmission using steered transmit diversity, as follows:

x _f(m)＝ E(m)· V(m)· S(m), formula (3)

Wherein the content of the first and second substances,s(m) is a number N_TX 1 vector having at most N to be transmitted in transmission span m_SA data symbol;

V(m) is N for transmission span m_T×N_TA steering matrix;

E(m) is a matrix of eigenvectors for transmission span m; and

x _f(m) is a number N_TX 1 vector with the data to be transmitted from N in the span m_TN transmitted by transmitting antenna_TA transmit symbol.

As shown in the formula (3),seach data symbol in (m) is composed ofVThe corresponding column of (m) is effectively spatially expanded. If N is present_S＜N_TThen use N_S×N_SMatrix arrayV(m) pairssN in (m)_SThe data symbols are spatially spread to obtain N_SAn "extension" symbol. Each spread symbol including N_SA portion of each of the data symbols. N to be spatially extended_SA spread symbol is sent toHN of (m)_SOn the eigenmode. Each steering matrixV(m) is a unitary matrix and may be generated as follows.

The receiving entity obtains the information from N_RReceived symbols of a receiving antenna, which can be usedExpressed as:

r _f(m)＝ H(m)· x _f(m)+ n(m)＝ H(m)· E(m)· V(m)· s(m)+ n(m), formula (4)

＝ H _{f_eff}(m)· s(m)+ n(m)

Wherein the content of the first and second substances,r _f(m) is a number N_RX 1 vector having N in transmission span m_RN obtained by one receiving antenna_RA received symbol;

n(m) is a noise vector for transmission span m;

H _{f_eff}(m) is a vector of datas(m) N observed for full CSI transmission using steered transmit diversity_R×N_TAn "effective" MIMO channel response matrix, which is:

H _{f_eff}(m)＝ H(m)· E(m)· V(m) of the reaction mixture. Formula (5)

For simplicity, assume that the noise is Additive White Gaussian Noise (AWGN) with a zero mean vector and a covariance matrix* _nn＝σ²· IWhere σ is²Is the variance of the noise andIis an identity matrix.

The receiving entity can recover using various receiver processing techniquessData symbols in (m). Techniques suitable for full CSI transmission include full CSI techniques and Minimum Mean Square Error (MMSE) techniques.

For full CSI techniques, receiveThe entity can derive a spatial filter matrixM _fcsi(m), as follows:

M _fcsi(m)＝ V ^H(m)· Λ ^-1(m)· E ^H(m)· H ^H(m) of the reaction mixture. Formula (6)

The receiving entity can useM _fcsi(m) performing receiver spatial processing as follows:

img id="idf0003" file="A20058002893900131.GIF" wi="191" he="21" img-content="drawing" img-format="GIF"/

img id="idf0004" file="A20058002893900132.GIF" wi="502" he="22" img-content="drawing" img-format="GIF"/

img id="idf0005" file="A20058002893900133.GIF" wi="162" he="24" img-content="drawing" img-format="GIF"/

formula (7)

Wherein the content of the first and second substances,is provided with N_SN of detected data symbols_TA x 1 vector; and

n _f(m) is post-detection noise after receiver spatial processing;

for MMSE techniques, the receiving entity may derive a spatial filter matrixM _{f_mmse}(m) is as follows：

img id="idf0007" file="A20058002893900142.GIF" wi="468" he="24" img-content="drawing" img-format="GIF"/Formula (8)

Spatial filter matrixM _{f_mmse}(m) estimating and summing symbols from the spatial filtersThe mean square error between the data symbols in (m) is minimized.

The receiving entity may perform MMSE spatial processing as follows:

img id="idf0008" file="A20058002893900143.GIF" wi="368" he="21" img-content="drawing" img-format="GIF"/

img id="idf0009" file="A20058002893900144.GIF" wi="383" he="22" img-content="drawing" img-format="GIF"/formula (9)

img id="idf0010" file="A20058002893900145.GIF" wi="413" he="23" img-content="drawing" img-format="GIF"/

Wherein the content of the first and second substances,D _{f_mmse}(m) is composed ofM _{f_mmse}(m)· H _{f_eff}(m) a diagonal matrix of diagonal elements,

alternatively, the first and second electrodes may be,D _{f_mmse}(m)＝diag[ M _{f_mmse}(m)· H _{f_eff}(m)](ii) a And

n _{f_mmse}(m) is MMSE filtered noise.

From spatial filtersM _{f_mmse}The symbol estimate of (m) is a non-normalized estimate of the data symbols. And scaling matrixD _{f_mmse}(m) the multiplication provides a normalized estimate of the data symbols.

Full CSI transmission attempts atH(m) data is transmitted on the eigenmodes. However, full CSI data transmission may not be completeOrthogonality due to, for example, imperfectionsH(m), errors in eigenvalue decomposition, limited arithmetic accuracy, etc. The MMSE technique can account for (or "clean up") any loss of orthogonality in the full CSI data transmission.

Table 1 summarizes the spatial processing for full CSI transmission using steered transmit diversity at the transmitting and receiving entities.

TABLE 1

2. Non-aligned mode-partial CSI transmission

For partial CSI transmission using pilot transmit diversity, the transmitting entity performs spatial processing as follows:

x _p(m)＝ V(m)· s(m), formula (10)

Wherein the content of the first and second substances,x _p(m) is the transmit data vector for transmission span m. As shown in the formula (10),seach data symbol in (m) is composed ofVThe corresponding column of (m) is spatially expanded. Then, from N_TA transmitting antenna for transmitting andV(m) N produced by multiplication_TAnd a spreading symbol.

The receiving entity obtains the received symbols, which can be expressed as:

r _p(m)＝ H(m)· x _p(m)+ n(m)＝ H(m)· V(m)· s(m)+ n(m), formula (11)

＝ H _{p_eff}(m)· s(m)+ n(m)，

Wherein the content of the first and second substances,r _p(m) is a vector of received symbols for transmission span m; and

H _{p_eff}(m) is composed ofs(m) N observed for partial CSI transmission using steered transmit diversity_R×N_TAn effective MIMO channel response matrix, which is:

H _{p_eff}(m)＝ H(m)· V(m) of the reaction mixture. Formula (12)

The receiving entity can recover using various receiver processing techniquessData symbols in (m). Techniques suitable for partial CSI transmission include Channel Correlation Matrix Inversion (CCMI) techniques (which are also commonly referred to as zero forcing techniques), MMSE techniques, and Successive Interference Cancellation (SIC) techniques.

For CCMI techniques, the receiving entity may derive a spatial filter matrixM _ccmi(m), as follows:

img id="idf0013" file="A20058002893900161.GIF" wi="525" he="24" img-content="drawing" img-format="GIF"/formula (13)

The receiving entity may perform CCMI spatial processing as follows:

img id="idf0014" file="A20058002893900162.GIF" wi="196" he="20" img-content="drawing" img-format="GIF"/

img id="idf0015" file="A20058002893900163.GIF" wi="358" he="23" img-content="drawing" img-format="GIF"/formula (14)

img id="idf0016" file="A20058002893900164.GIF" wi="130" he="16" img-content="drawing" img-format="GIF"/

Wherein the content of the first and second substances,n _ccmi(m) is CCMI filter noise. Due to the fact thatR _{p_eff}With the structure of (m), the CCMI technique may amplify noise.

For MMSE techniques, the receiving entity may derive a spatial filter matrixM _{p_mmse}(m), as follows:

img id="idf0017" file="A20058002893900165.GIF" wi="421" he="22" img-content="drawing" img-format="GIF"/formula (15)

Equation (15) for partial CSI transmission has the same form as equation (8) for full CSI transmission. However, it is used in formula (15) for partial CSI transmissionH _{p_eff}(m) (instead ofH _{f_eff}(m))。

The receiving entity may perform MMSE spatial processing as follows:

img id="idf0018" file="A20058002893900166.GIF" wi="326" he="23" img-content="drawing" img-format="GIF"/formula (16)

img id="idf0019" file="A20058002893900167.GIF" wi="411" he="22" img-content="drawing" img-format="GIF"/

Wherein the content of the first and second substances,D _{p_mmse}(m)=diag[ M _{p_mmse}(m)· H _{p_eff}(m)]andn _{p_mmse}(m) is the MMSE filtering noise for partial CSI transmission.

For SCI techniques, the transmitting entity recovers in successive stagessData symbols in (m). For clarity, the following description assumess(m) elements andr _peach element of (m) corresponds to a data symbol stream. Receiving entity is in N_STreated in one continuous stager _pN in (m)_RA stream of received symbols, thereby recoveringsN in (m)_SA stream of data symbols. Generally, SIC processing results in one packet being recovered for one flow, then another packet being recovered for another flow, and so on. For simplicity, the following description assumes N_S=N_T。

In each stage l, wherein l 1_SFor which the receiving entity is paired with N_RA stream of input symbolsr _p ^l(m) performing receiver spatial processing. The input symbol stream for the first stage (l ═ 1) is a received symbol stream, orimg id="idf0020" file="A20058002893900171.GIF" wi="115" he="22" img-content="drawing" img-format="GIF"/For each subsequent stage (l ═ 2.. N)_S) Is a modified symbol stream of a previous stage. Receiver spatial processing for stage l is based on a spatial filter matrixM _x ^l(m) that may be based on a reduced effective channel response matrixH _{p_eff} ^l(m) and further derived from CCMI, MMSE, or other techniques.H _{p_eff} ^l(m) including and not yet recovered in stage lComplex N_S-l +1 data symbol streams correspondH _{p_eff}N in (m)_SL +1 columns. The receiving entity obtains a stream of detected data symbols for phase lAnd further processes (e.g., demodulates, deinterleaves, and decodes) the stream to obtain a corresponding decoded data stream

The receiving entity then estimates the data symbol stream s_lCause interference that is not yet recovered by other data symbol streams. To estimate the interference, the receiving entity processes (e.g., re-encodes, interleaves, and symbol maps) the decoded data stream in the same manner as was performed for that stream by the transmitting entityAnd obtain a stream of "remodulated" symbolsWhich is the stream of just recovered data symbols s_lAnd (4) estimating. The receiving entity then uses the steering matrixV(m) spatially processing the re-modulated symbol stream and further correlating the result with a channel response matrixH(m) to obtain a composed stream s_lCaused by N_RAn interference componenti ^l(m) of the reaction mixture. The receiving entity then derives N for the current phase l from_RA stream of input symbolsr _p ^l(m) subtracting N_RAn interference componenti ^l(m) to obtain N for the next stage_RA stream of input symbolsr _p ^l+1(m) orimg id="idf0025" file="A20058002893900176.GIF" wi="182" he="31" img-content="drawing" img-format="GIF"/Input symbol streamr _p ^l+1(m) indicates if the data symbol stream s is not transmitted_lAnd the receiving entity will receive the stream, assuming interference cancellation is performed efficiently. Then, the receiving entity pair N_RA stream of input symbolsr _p ^l+1(m) repeating the same process to recover another data stream. However, the effective channel response matrix for the subsequent phase l +1H _{p_eff} ^l+1(m) is subtracted from the data symbol stream s recovered in stage l_lA corresponding column.

For the SIC technique, the SNR of each data symbol stream depends on: (1) the receiver processing technique used for each stage (e.g., CCMI or MMSE), (2) the particular stage in which the data symbol stream is recovered, and (3) the amount of interference, since the data symbol stream is recovered in a later stage. Generally, the SNR gradually improves the data symbol stream recovered in the later stage because the interference of the data symbol stream recovered in the previous stage is cancelled. This can then use the higher rate for the data symbol stream recovered in a later stage.

Table 2 summarizes the spatial processing for partial CSI transmission using steered transmit diversity at the transmitting and receiving entities. For simplicity, the SIC technique is not shown in table 2.

TABLE 2

Fig. 2 shows a model for data transmission using steered transmit diversity. The transmitting entity 110 performs spatial processing (or spatial spreading) for pilot transmit diversity (block 220) and spatial processing for full CSI or partial CSI transmission (block 230). The receiving entity 150 performs receiver spatial processing for full CSI or partial CSI transmission (block 260) and receiver spatial processing (or spatial despreading) for pilot transmit diversity (block 270). As shown in fig. 2, the transmitting entity performs spatial spreading for steered transmit diversity before spatial processing (if any) for full-CSI and partial-CSI transmissions. The receiving entity may perform complementary receiver spatial processing for full-CSI or partial-CSI transmission prior to spatial despreading for pilot transmit diversity.

3. Spatial filter matrix computation

Using pilot transmit diversity, different pilot matricesV(m) may be used for different transmission spans to randomize the effective MIMO channel observed by the data transmission. This may then improve performance because data transmission does not observe "poor" MIMO channel realizations for an extended period of time. The transmission span may correspond to a symbol period for a single carrier MIMO system or a subband for a multicarrier MIMO system.

Fig. 3A illustrates partial CSI transmission using steered transmit diversity for a single carrier MIMO system. For this system, the transmission span index m may be equal to the symbol period index n (or m ═ n). A vector of data symbols may be applied in each symbol period ns(n) transmitting and using the steering matrix selected for the symbol periodV(n) pairss(n) performing spatial expansion. Each data symbol vectors(n) observation ofH _{p_eff}(n)＝ H(n)· V(n) effective MIMO channel response and use of spatial filter matrixM _x(n) recoverys(n)。

Fig. 3B illustrates partial CSI transmission using steered transmit diversity in a multi-carrier MIMO system. For this system, the transmission span index m may be equal to the sub-band index k (or m-k). For each symbol period, a vector of data symbols may be transmitted in each subband ks(k) And using steering matrices selected for that subbandV(k) Will be provided withs(k) And performing space expansion. Each data symbol vectors(k) Observe thatH _{p_eff}(k)＝ H(k)· V(k) And using a spatial filter matrixM _x(k) Will be provided withs(k) And (6) recovering. Vector quantitys(k) Sum matrixV(k)、H(k) AndM _x(k) also a function of the symbol period n, but is not shown for simplicity.

As shown in fig. 3A and 3B, the spatial filter matrix used by the receiving entity is a function of the transmission span index m if different steering matrices are used for different transmission spans. Even if the channel response matrixHIt is also true that (m) is fixed or constant over a series of transmission spans. In a multi-carrier MIMO system, for example, for a flat fading MIMO channel with a flat frequency response,H(k) may be fixed across the subband groups. As another example, in a single carrier MIMO system, for a MIMO channel without time fading,H(n) may be fixed for a given time interval. The time interval may correspond to all or a portion of the duration for transmitting a block of data symbols that is encoded and decoded in blocks.

The correlation typically exists between channel response matrices for adjacent transmission spans, e.g. betweenH(m) andH(m +/-1). The correlation may be used to simplify the computation of the spatial filter matrix at the receiving entity. The following is directed toThe calculation of one case-full correlation and partial correlation-is described.

A. All correlation

With full correlation, the channel response matrix for a MIMO channel is fixed within a series of correlated transmission span indices, e.g., M1. Therefore, the temperature of the molten metal is controlled,H(1)＝ H(2)＝...＝ H(M)＝ H。

for full-CSI techniques, a spatial filter matrix using a fully-correlated channel response matrix may be usedM _fcsi(m) is represented by:

M _fcsi(m)＝ V ^H(m)· Λ ^-1(m)· E ^H(m)· H ^H. Formula (17)

The spatial filter matrix may then be appliedM _fcsi(m) is calculated as:

M _fcsi(m)＝ V ^H(m)· M _{fcsi_base}m, where M1.. M, formula (18)

Wherein the content of the first and second substances,M _{fcsi_base}＝ Λ ^-1· E ^H· H ^His a basic spatial filter matrix, i.e., a spatial filter matrix for full CSI techniques without using pilot transmit diversity. Basic spatial filter matrixM _{fcsi_base}Is not a function of the transmission span m because of the channel response matrixHIs stationary. Equation (18) represents the spatial filter matrix for each transmission span mM _fcsi(m) the filter matrix can be formed by combining a basic spatial filter matrix M _{fcsi_base}Pre-multiplying the steering matrix for the transmission spanV ^H(m) to obtain.

Alternatively, the spatial filter matrix may be usedM _fcsi(m) is calculated as:

M _fcsi(m)＝ W ₁(m)· M _fcsi(1) m, where M is 2.. M, formula (19)

Wherein the content of the first and second substances,M _fcsi(1)＝ V ^H(1)· Λ ^-1· E ^H· H ^HandW ₁(m)＝ V ^H(m)· V(1). Formula (19)

Representing a spatial filter matrix for each transmission span mM _fcsi(m) the spatial filter matrix can be used for transmission span 1M _ccmi(1) Pre-multiplying by a matrixW ₁(m) to obtain. Matrix arrayW ₁(M) are unitary matrices, where M2V(m) andV(1) to be acquired. Can be combined with matrixW ₁(m) are pre-calculated and stored in memory.

For MMSE techniques for full CSI transmission, a spatial filter matrix using a fully correlated channel response matrix may be usedM _{f_mmse}(m) is represented by:

img id="idf0027" file="A20058002893900211.GIF" wi="456" he="32" img-content="drawing" img-format="GIF"/

img id="idf0028" file="A20058002893900212.GIF" wi="436" he="22" img-content="drawing" img-format="GIF"/formula (20)

img id="idf0029" file="A20058002893900213.GIF" wi="318" he="21" img-content="drawing" img-format="GIF"/

Equation (20) is the usage propertyA· B)^-1＝ B ^-1· A ^-1AndV· V ^H＝ Iand (4) deriving. The term in brackets in the second equation in equation (20) can be expressedExpressed as:

[ V ^H· E ^H· H ^H· H· E· V+σ²· I]＝[ V ^H( E ^H· H ^H· H· E+σ²· V· I· V ^H)· V]，＝[ V ^H( E ^H· H ^H· H· E+σ²· I)· V]，

wherein "(m)" is omitted for clarity. The inverse of the term in the second equation above can then be expressed as:

[ V ^H( E ^H· H ^H· H· E+σ²· I)· V]^-1＝[ V ^H( E ^H· H ^H· H· E+σ²· I)^-1· V]，

wherein the content of the first and second substances,V ^H＝ V ^-1。

the spatial filter matrix can be formedM _{f_mmse}(m) is calculated as:

M _{f_mmse}(m)＝ V ^H(m)· M _{f_mmse_base}m, where M1.. M, formula (21)

Wherein the content of the first and second substances,M _{f_mmse_base}＝[ E ^H· H ^H· H· E+σ²· I]^-1· E ^H· H ^H. Similar to the full-CSI technique, the spatial filter matrix for the transmission span mM _{f_mmse}(m) the filter matrix can be formed by combining a basic spatial filter matrixM _{f_mmse_base}(m) pre-multiplying by steering matrixV ^H(m) to obtain. The spatial filter matrix can also be combinedM _{f_mmse}(m) is calculated as:

M _{f_mmse}(m)＝ W ₁(m)· M _{f_mmse}(1) m, where M2.. M, formula (22)

Wherein the content of the first and second substances,M _{f_mmse}(1)＝ V ^H(1)·[ E ^H· H ^H· H· E+σ²· I]^-1· E ^H· H ^H。

for CCMI techniques, a spatial filter matrix using a fully correlated channel response matrix may be usedM _ccmi(m)Expressed as:

img id="idf0030" file="A20058002893900214.GIF" wi="344" he="24" img-content="drawing" img-format="GIF"/

img id="idf0031" file="A20058002893900215.GIF" wi="294" he="23" img-content="drawing" img-format="GIF"/

img id="idf0032" file="A20058002893900216.GIF" wi="258" he="22" img-content="drawing" img-format="GIF"/formula (23)

img id="idf0033" file="A20058002893900217.GIF" wi="282" he="23" img-content="drawing" img-format="GIF"/

img id="idf0034" file="A20058002893900218.GIF" wi="140" he="21" img-content="drawing" img-format="GIF"/

Wherein, the [ alpha ], [ beta ]V ^H(m)]^-1＝ V(m) becauseV(m) is a unitary matrix.

Thus, a spatial filter matrix can be formedM _ccmi(m) is calculated as:

M _ccmi(m)＝ V ^H(m)· M _{ccmi_base}m, where M1.. M, formula (24)

Wherein the content of the first and second substances,M _{ccmi_base}＝ R ^-1· H ^H. The spatial filter matrix can also be combinedM _ccmi(m) is calculated as:

M _ccmi(m)＝ W ₁(m)· M _ccmi(1) m, where M is 2.. M, formula (25)

Wherein the content of the first and second substances,M _ccmi(1)＝ V ^H(1)· R ^-1· H ^H。

for MMSE techniques for partial CSI transmission, a spatial filter matrix using a fully correlated channel response matrix may be usedM _{p_mmse}(m) is represented by:

img id="idf0035" file="A20058002893900221.GIF" wi="415" he="23" img-content="drawing" img-format="GIF"/

img id="idf0036" file="A20058002893900222.GIF" wi="348" he="20" img-content="drawing" img-format="GIF"/formula (26)

img id="idf0037" file="A20058002893900223.GIF" wi="230" he="20" img-content="drawing" img-format="GIF"/

Equation (26) may be derived in a similar manner as equation (20) above.

The spatial filter matrix can be formedM _{p_mmse}(m) is calculated as:

M _{p_mmse}(m)＝ V ^H(m)· M _{p_mmse_base}m, where M1.. M, formula (27)

Wherein the content of the first and second substances,M _{p_mmse_base}＝[ H ^H· H+σ²· I]^-1· H ^H. The spatial filter matrix can also be combinedM _{p_mmse}(m) is calculated as:

M _{p_mmse}(m)＝ W ₁(m)· M _{p_mmse}(1) m, where M2.. M, formula (28)

Wherein the content of the first and second substances,M _{p_mmse}(1)＝ V ^H(1)·[ H ^H· H+σ²· I]^-1· H ^H。

table 3 summarizes the calculations of spatial filter matrices for full-CSI and partial-CSI transmissions using fully correlated channel response matrices in transmission span M1.

TABLE 3 spatial Filter matrix Using full correlation

Mode(s)	Spatial filter matrix	Technique of
			Full CSI	M fcsi_base＝ Λ -1· E H· H HAnd anM fcsi(m)＝ V H(m)· M fcsi_base	Full CSI
M f_mmse_base＝[ E H· H H· H· E+σ2· I]-1· E H· H HAnd anM f_mmse_base(m)＝ V H(m)· M f_mmse_base	MMSE
		Partial CSI		M ccmi_base＝ R -1· H HAnd anM ccmi(m)＝ V H(m)· M ccmi_base	CCMI

M p_mmse_base＝[ H H· H+σ2· I]-1· H HAnd anM p_mmse(m)＝ V H(m)· M p_mmse_base

MMSE

In general, the spatial filter matrix for the transmission span m can be calculated asM _x(m)＝ V ^H(m)· M _{x_base}Where the subscript "x" denotes the receiver processing technology and may be "fcsi", "f _ mmse", "ccmi", or "p _ mmse". The basic spatial filter matrix can also be computed if no pilot transmit diversity is usedM _{x_base}。

Fig. 4 shows a flow diagram of the calculation by processor 400 of a spatial filter matrix using a fully correlated channel response matrix in transmission span M1. First, an initial spatial filter matrix is calculatedM _{x_init}(block 412). The initial spatial filter matrix may be a basic spatial filter matrixM _{x_basse}It is derived based on: (1) channel response matrixH(ii) a And (2) the receiver processing technique selected for use (e.g., full CSI, MMSE for full CSI, CCMI, or MMSE for partial CSI). Of course, the initial spatial filter matrix may also be a spatial filter matrix for a transmission span m-1M _x(1) Which may be based onHAndV(1) and (6) exporting.

Then, ifM _{x_init}＝ M _{x_base}The transmission span index m is set to 1 (as shown in fig. 4), or, if so, ifM _{x_init}＝ M _x(1) It is set to 2 (block 414). Then, based on the initial spatial filter matrixM _{x_init}And a steering matrix for the transmission span mV(m) calculating a spatial filter matrix for a transmission span mM _x(m) (block 416). In particular, can be based onM _{x_base}AndV(m) orM _x(1) AndW ₁(m) calculatingM _x(m) as described above. A determination is then made as to whether M < M (block 420). If the answer is "yes," then the metric m is incremented (block 422) and the process returns to block 416 to compute a spatial filter matrix for another transmission span. Otherwise, if M is M in block 420, the spatial filter matrix isM _x(1) ToM _x(nm) for receiving symbol vectors, respectivelyr _x(1) To r _xReceiver spatial processing of (M) (block 424). Although not shown in FIG. 4 for simplicity, once the spatial filter matrix is generatedM _x(m) and obtaining a vector of received symbolsr _x(m), then each spatial filter matrix can be used for receiver spatial processing.

For full CSI transmission, the spatial processing at the transmitting entity may also be simplified to:x _f(m)＝ E· V(m)· s(m) of the reaction mixture. May be based on steering matrices for each transmission span mV(m) and matrixECalculating a matrix for the transmission spanE· V(m)， EIs not a function of the transmission span for the fully correlated case.

B. Partial correlation

With partial correlation, the channel response matrix for the MIMO channel is not fully correlated over a series of correlated transmission span indices. In this case, the spatial filter matrix calculated for the transmission span l may be used to help calculate a spatial filter matrix for another transmission span m.

In one embodiment, the steering matrix for transmission span l is removed by cullingV(l) From a spatial filter matrix calculated for the transmission span lM _x(l) To obtain a basic spatial filter matrix for a transmission span lM _{x_base}(l) The following are:

M _{x_base}(l)＝ V(l)· M _x(l) In that respect Formula (29)

Then, the basic spatial filter matrixM _{x_base}(l) Method for deriving a basic spatial filter matrix for a transmission span m (e.g. m ═ l +1) M _{x_base}(m) of the reaction mixture. For example, repeated pairs may be usedM _{x_base}(l) Performing an iterative process or operation of a set of calculations to calculateM _{x_base}(m) to obtainM _{x_base}(m) final scheme. Iterative processes (e.g., adaptive MMSE operations, gradient operations, lattice operations, etc.) for computing an MMSE scheme are known in the art and are not described herein. Spatial filter matrix that can be used for transmission span mM _x(m) is calculated as:

M _x(m)＝ V ^H(m)· M _{x_base}(m) of the reaction mixture. Formula (30)

Therefore, the processing order of this embodiment can be given as:M _x(l)→ M _{x_base}(l)* M _{x_base}(m)→ M _x(m) where "→" denotes direct calculation and "*" denotes possible iterative calculation. Basic spatial filter matrixM _{x_base}(l) AndM _{x_base}(m) excluding steering matrices and spatial filter matricesM _x(l) AndM _x(m) steering matrices for transmission spans l and m, respectivelyV(l) AndV(m)。

in another embodiment, initial guessing is usedThe iterative process of performing a set of calculations is repeated to calculate a spatial filter matrix for a transmission span mM _x(m) of the reaction mixture. Can be derived from a spatial filter matrix derived for the transmission span lM _x(l) To derive an initial guess, as follows:

img id="idf0039" file="A20058002893900251.GIF" wi="185" he="20" img-content="drawing" img-format="GIF"/formula (31)

Wherein the content of the first and second substances,W ₁(m)＝ V _H(m)· V(l) In that respect The processing order for this embodiment can be given as:img id="idf0040" file="A20058002893900252.GIF" wi="210" he="20" img-content="drawing" img-format="GIF"/spatial filter matrixAndM _x(m) each comprise steering matrices for a transmission span mV(m)。

For the above embodiments, theM _{x_base}(l) Andis considered to be used to deriveSpatial filter matrix for new transmission span mM _x(m) an initial spatial filter matrix. In general,M _x(l) AndM _xthe amount of correlation between (m) depends onM _{x_base}(l) AndM _{x_base}(m) amount of correlation between, which depends on the transmission spans l and mH(l) AndH(m) relative amount between (m). Higher correlation may result in faster convergence to useM _x(l) The final scheme of (1).

Fig. 5 shows a flow diagram of a process 500 for spatial filter matrix calculation using a spatial filter matrix of a partial correlation channel response matrix for a transmission span M1. The metrics for the current and next transmission spans are initialized to l-1 and m-2 (block 512). Calculating a spatial filter matrix for a transmission span l based on selected receiver processing techniquesM _x(l) (block 514). Then, based on the spatial filter matrixM _x(l) And a suitable steering matrixV(l) AndV(m) to compute an initial spatial filter matrix for the transmission span mM _{x_init}E.g., as shown in equations (29) or (31) (block 516). Then, for example, using an iterative process, based on the initial spatial filter matrixM _{x_init}To calculate a spatial filter matrix for a transmission span mM _x(m) (block 518).

A determination is then made as to whether M < M (block 520). If the answer is "yes," the indicators l and m are updated, e.g., to l ═ m and m ═ m +1 (block 522). The process then returns to block 516 to calculate a spatial filter matrix for another transmission span. Otherwise, if all spatial filter matrices have been calculated, as determined in block 520, the spatial filter matrices are calculatedM _x(1) ToM _x(M) for receiving symbol vectors, respectively r _x(1) Tor _xReceiver spatial processing of (M) (block 524).

For simplicity, fig. 5 shows the computation of M spatial filter matrices for M consecutive transmission spans M-1.. M. The transmission spans need not be contiguous. Typically, the derived spatial filter matrix for one transmission span l is used to obtain an initial guess of the spatial filter matrix for another transmission span m, where l and m may be any index values.

4. Steering matrix

A set of steering matrices (or transmit matrices) may be generated and used for steering transmit diversity. These steering matrices can be represented asVEither alone orV(i) L, where L is any integer value greater than 1. Each steering matrixV(i) Should be unitary. This condition ensures the utilization ofV(i) N of simultaneous emission_TThe data symbols have the same power and are being utilizedV(i) After spatial spreading, are orthogonal to each other.

The set of L steering matrices may be generated in various ways. For example, L steering matrices may be generated based on a unitary base matrix and a set of scalars. The basic matrix may be used as one of the L steering matrices. The other L-1 steering matrices may be generated by multiplying the rows of the base matrix by different combinations of scalars. Each scalar may be any real or complex value. The scalars are selected to have unity magnitude so that the steering matrix generated using these scalars is unitary.

The base matrix may be a Walsh matrix. A 2 x 2 Walsh matrixW _2×2And a larger size Walsh matrixW _2N×2NCan be expressed as follows:

img id="idf0043" file="A20058002893900261.GIF" wi="103" he="47" img-content="drawing" img-format="GIF"/andimg id="idf0044" file="A20058002893900262.GIF" wi="206" he="49" img-content="drawing" img-format="GIF"/formula (32)

The dimensions of the Walsh matrix are powers of 2 (e.g., 2, 4, 8, etc.).

The base matrix may also be a fourier matrix. For NxN Fourier matrixD _N×N， D _N×NRow n and column m of element d_m，nCan be expressed as:

img id="idf0045" file="A20058002893900263.GIF" wi="144" he="33" img-content="drawing" img-format="GIF"/wherein N ═ 1.. N } and m ═ 1.. N }. Formula (33)

A fourier matrix of arbitrary square dimensions (e.g., 2, 3, 4, 5, etc.) may be formed. Other matrices may also be used as the base matrix.

For an nxn base matrix, each of rows 2 through N of the base matrix may be independently multiplied by one of K different possible scalars. K from K scalars for N-1 lines^N-1In different substitution forms, to obtain K^N-1A different steering matrix. For example, each of rows 2 through N is independently multiplied by a scalar +1, -1, + j, or-j. For N-4 and K-4, with four different scalars, 64 different steering matrices can be generated from the 4 × 4 basis matrix. In general, each row of the base matrix may be multiplied by having e^jθAny scalar of the form where θ can be any phase value. For each element of an N x N basic matrix multiplied by a scalar quantityFurther scaling to obtain an nxn steering matrix per column per unit power.

Steering matrices derived based on Walsh matrices (i.e., 4 x 4 fourier matrices) have certain desirable properties. If the rows of the Walsh matrix are multiplied by scalars ± 1 and ± j, each element of the resulting steering matrix is +1, -1, + j, or-j. In this case, one element of the spatial filter matrix (i.e., the "weight") can be multiplied by one element of the steering matrix by only a bit operation. The computation for deriving the spatial filter matrix for the full correlation case can be greatly simplified if the elements of the L steering matrices belong to the set consisting of { +1, -1, + j, -j.

5. MIMO system

Fig. 6 shows a block diagram of an access point 610 and a user terminal 650 in a MIMO system 600. Access point 610 is equipped with N_apAn antenna for data transmission and reception, and a user terminal equipped with N_utAn antenna, wherein N_ap> 1 and N_ut＞1。

On the downlink, at access point 610, a TX data processor 620 receives and processes (encodes, interleaves, and symbol maps) traffic/packet data and control/overhead data and provides data symbols. TX spatial processor 630 utilizes steering matricesV(m) and possibly eigenvector matrices for the downlinkE(m) spatially processing the data symbols, e.g., as shown in tables 1 and 2. TX spatial processor 630 also multiplexes using pilot symbols, as appropriate, and combines N_apOne transmit symbol stream is provided to N_apAnd transmitter units 632a through 632 ap. Each transmitter unit 632 receives and processes a respective transmit symbol stream and provides a respective downlink modulated signal. From N_apN from transmitter units 632a through 632ap are transmitted by antennas 634a through 634ap, respectively_apA downlink modulated signal.

At user terminal 650, N_utThe transmitted downlink modulated signals are received by antennas 652a through 652ut, which each provide a received signal to a respective receiver unit 654. Receiver units 654 complementarily process the processing performed by receiver units 632 and provide received symbols. RX spatial processor 660 for data from all N_utThe received symbols for each receiver unit 654a through 654ut are receiver spatially processed, e.g., as shown in tables 1 and 2, and provide detected data symbols. An RX data processor 670 processes (e.g., symbol demaps, deinterleaves, and decodes) the detected data symbols and provides decoded data for the downlink.

The processing for the uplink may be the same as or different from the processing for the downlink. Traffic and control data are processed (e.g., encoded, interleaved, and symbol mapped) by a TX data processor 688 and by a TX spatial processor 690 using steering matricesV(m) and possibly eigenvector matrices for the uplinkE(m) spatially processed and multiplexed with pilot symbols to generate N_utA stream of transmit symbols. N is a radical of_utA transmitter unit 654 a-654 ut for the N_utShaping the transmitted symbol streams to generate N_utAn uplink modulation signal via N_utThe antennas 652a through 652ut transmit.

At the access point 610, the uplink modulated signal consists of N_apA plurality of antennas 634a to 634ap for receiving and transmitting by N_apEach receiver unit 632a through 632ap processes to obtain received symbols for the uplink. An RX spatial processor 644 performs receiver spatial processing on the received symbols and provides detected data symbols, which are further processed by an RX data processor 646 to obtain decoded data for the uplink.

Processors 638 and 678 perform channel estimation and spatial filter matrix computation for the access point and user terminal, respectively. Controllers 640 and 680 control the operation of various processing units at the access point and user terminals, respectively. Memory units 642 and 682 store data and program codes used by controllers 630 and 680, respectively.

Fig. 7 illustrates an embodiment of a processor 678 that performs channel estimation and spatial filter matrix computation for user terminal 650. Channel estimator 712 obtains received pilot symbols and derives a channel response matrix for each transmission span in which the received pilot symbols are available. Filter 714 may time-domain filter the channel response matrices for the current and previous transmission spans to obtain a higher quality channel response matrixH(m) of the reaction mixture. Unit 716 then calculates an initial spatial filter matrixM _{x_init}。

For full correlationH(m), initial spatial filter matrixM _{x_init}Can be as follows: (1) based onH(m) and the basic spatial filter matrix calculated by the selected receiver processing techniqueM _{x_base}(ii) a Or (2) based onH(1)、 V(1) Spatial filter matrix for transmission span 1 calculated with selected receiver processing techniqueM _x(1). For partial correlationH(m), initial spatial filter matrixM _{x_init}May be an initial guessM _{x_base}(l) Or based on a calculated spatial filter matrix for another transmission span/M _x(l) Obtained byUnit 718 is based on an initial spatial filter matrixM _{x_init}And a steering matrix for the transmission span mV(m) calculating a spatial filter matrix for the transmission spanM _x(m) of the reaction mixture. For partial correlationH(m), unit 718 may be based on the initial spatial filter matrix pairM _x(m) performing a calculation of an iterative process which isM _x(m) initial guess.

Processor 638 performs channel estimation and spatial filter matrix computations for access point 610 and can be implemented in a similar manner as processor 678.

The spatial filter matrix computation techniques described herein may be implemented in various ways. For example, these techniques may be implemented in hardware, software, or a combination of hardware and software. For a hardware implementation, the processing units used for spatial filter matrix computation may be implemented within one or more Application Specific Integrated Circuits (ASICs), Digital Signal Processors (DSPs), Digital Signal Processing Devices (DSPDs), Programmable Logic Devices (PLDs), Field Programmable Gate Arrays (FPGAs), processors, controllers, micro-controllers, microprocessors, other electronic units designed to perform the functions described herein, or a combination thereof.

For a software implementation, the spatial filter matrix computation may be implemented with modules (e.g., procedures, functions, etc.). These software codes may be stored in memory units (e.g., memory units 642 and 682 in fig. 6) and executed by processors (e.g., controllers 640 and 680 in fig. 6). The memory unit may be implemented within the processor or external to the processor, in which case it can be communicatively coupled to the processor via various means as is known in the art.

Headings are included herein for reference, which are intended to locate specific sections. These headings are not intended to limit the scope of the concepts described therein under, which concepts may be applied to other sections throughout the specification.

The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims (38)

1. A method of deriving a spatial filter matrix in a wireless multiple-input multiple-output (MIMO) communication system, comprising:

determining an initial spatial filter matrix; and

deriving a plurality of spatial filter matrices for a plurality of transmission spans based on the initial spatial filter matrix and a plurality of steering matrices for the plurality of transmission spans.

2. The method of claim 1, wherein the initial spatial filter matrix is determined based on a channel response matrix for a MIMO channel.

3. The method of claim 2, wherein the spatial filter matrix for each of the plurality of transmission spans is derived based on the initial channel response matrix and a steering matrix for the transmission span.

4. The method of claim 2, wherein the initial spatial filter matrix is determined based further on a steering matrix for one of the plurality of transmission spans.

5. The method of claim 3, wherein the spatial filter matrix for each of the plurality of transmission spans is derived based on the initial channel response matrix, the steering matrix used to determine the initial spatial filter matrix, and a steering matrix for the transmission span.

6. The method of claim 1, wherein data is transmitted on orthogonal spatial channels of a MIMO channel, and wherein the initial channel response matrix is determined according to a full channel state information (full CSI) technique.

7. The method of claim 1, wherein data is transmitted on orthogonal spatial channels of a MIMO channel, and wherein the initial channel response matrix is determined according to a Minimum Mean Square Error (MMSE) technique.

8. The method of claim 1, wherein data is transmitted on spatial channels of a MIMO channel, and wherein the initial channel response matrix is determined according to a Channel Correlation Matrix Inversion (CCMI) technique.

9. The method of claim 1, wherein data is transmitted on a spatial channel of a MIMO channel, and wherein the initial channel response matrix is determined according to a Minimum Mean Square Error (MMSE) technique.

10. The method of claim 1, wherein a transmitting entity performs spatial processing on data using the plurality of steering matrices to achieve transmit diversity.

11. The method of claim 1, wherein elements of the plurality of steering matrices are members of the set consisting of +1, -1, + j, and-j, where j is the square root of-1.

12. The method of claim 1, wherein the plurality of transmission spans correspond to a plurality of symbol periods.

13. The method of claim 1, wherein the plurality of transmission spans correspond to a plurality of frequency subbands.

14. The method of claim 1, further comprising:

spatially processing symbols received over the plurality of transmission spans using the plurality of spatial filter matrices.

15. An apparatus in a wireless multiple-input multiple-output (MIMO) communication system, comprising:

a processor that determines an initial spatial filter matrix and derives a plurality of spatial filter matrices for a plurality of transmission spans based on the initial spatial filter matrix and a plurality of steering matrices for the plurality of transmission spans; and

a memory storing the plurality of steering matrices.

16. The apparatus of claim 15, wherein the initial spatial filter matrix is determined based on a channel response matrix for a MIMO channel, and wherein the spatial filter matrix for each of the plurality of transmission spans is derived based on the initial channel response matrix and a steering matrix for the transmission span.

17. The apparatus of claim 15, wherein the initial channel response matrix is determined according to a full channel state information (full CSI) technique, a Minimum Mean Square Error (MMSE) technique, or a Channel Correlation Matrix Inversion (CCMI) technique.

18. The apparatus of claim 15, wherein elements of the plurality of steering matrices are members of the set consisting of +1, -1, + j, and + j, where j is the square root of-1.

19. The apparatus of claim 15, further comprising:

a spatial processor that spatially processes symbols received over the plurality of transmission spans using the plurality of spatial filter matrices.

20. An apparatus in a wireless multiple-input multiple-output (MIMO) communication system, comprising:

a determining module that determines an initial spatial filter matrix; and

a derivation module that derives a plurality of spatial filter matrices for a plurality of transmission spans based on the initial spatial filter matrix and a plurality of steering matrices for the plurality of transmission spans.

21. The apparatus of claim 20, wherein the initial spatial filter matrix is determined based on a channel response matrix for a MIMO channel, and wherein the spatial filter matrix for each of the plurality of transmission spans is derived based on the initial channel response matrix and a steering matrix for the transmission span.

22. The apparatus of claim 20, wherein the initial channel response matrix is determined according to a full channel state information (full CSI) technique, a Minimum Mean Square Error (MMSE) technique, or a Channel Correlation Matrix Inversion (CCMI) technique.

23. The apparatus of claim 20, wherein elements of the plurality of steering matrices are members of the set consisting of +1, -1, + j, and + j, where j is the square root of-1.

24. The apparatus of claim 20, further comprising:

a spatial processing module to spatially process symbols received over the plurality of transmission spans using the plurality of spatial filter matrices.

25. A method of deriving a spatial filter matrix in a wireless multiple-input multiple-output (MIMO) communication system, comprising:

deriving a first spatial filter matrix for the first transmission span;

determining a first initial spatial filter matrix for a second transmission span based on the first spatial filter matrix; and

deriving a second spatial filter matrix for the second transmission span based on the first initial spatial filter matrix.

26. The method of claim 25, wherein the first spatial filter matrix is derived based on a channel response matrix obtained for a MIMO channel in the first transmission span and further based on a receiver spatial processing technique.

27. The method of claim 25, wherein the determining a first initial spatial filter matrix comprises:

processing the first spatial filter matrix to reject a first steering matrix for the first transmission span, and wherein the first initial spatial filter matrix is equal to the first spatial filter matrix from which the first steering matrix is rejected.

28. The method of claim 25, wherein the determining a first initial spatial filter matrix comprises:

processing the first spatial filter matrix to reject a first steering matrix for the first transmission span and to include a second steering matrix for the second transmission span, and wherein the first initial spatial filter matrix is equal to the first spatial filter matrix with the first steering matrix rejected and the second steering matrix included.

29. The method of claim 25, wherein the second spatial filter matrix is derived using an iterative process that iteratively performs a set of calculations on the first initial spatial filter matrix to obtain a final solution for the second spatial filter matrix.

30. The method of claim 25, further comprising:

determining a second initial spatial filter matrix for a third transmission span based on the second spatial filter matrix; and

deriving a third spatial filter matrix for the third transmission span based on the second initial spatial filter matrix.

31. The method of claim 25, wherein the first and second transmission spans correspond to two different symbol periods.

32. The method of claim 25, wherein the first and second transmission spans correspond to two different frequency subbands.

33. An apparatus in a wireless multiple-input multiple-output (MIMO) communication system, comprising:

a processor that derives a first spatial filter matrix for a first transmission span, determines a first initial spatial filter matrix for a second transmission span based on the first spatial filter matrix, and derives a second spatial filter matrix for the second transmission span based on the first initial spatial filter matrix.

34. The apparatus of claim 33, wherein the processor processes the first spatial filter matrix to cull a first steering matrix for the first transmission span, and wherein the first initial spatial filter matrix is equal to the first spatial filter matrix culled the first steering matrix.

35. The apparatus of claim 33, wherein the processor further determines a second initial spatial filter matrix for a third transmission span based on the second spatial filter matrix and derives a third spatial filter matrix for the third transmission span based on the second initial spatial filter matrix.

36. An apparatus in a wireless multiple-input multiple-output (MIMO) communication system, comprising:

a first spatial filter matrix derivation module to derive a first spatial filter matrix for a first transmission span;

a first initial spatial filter matrix determination module that determines a first initial spatial filter matrix for a second transmission span based on the first spatial filter matrix; and

a second spatial filter matrix derivation module to derive a second spatial filter matrix for the second transmission span based on the first initial spatial filter matrix.

37. The apparatus of claim 36, wherein the first initial spatial filter matrix determination module comprises:

a processing module that processes the first spatial filter matrix to reject a first steering matrix for the first transmission span, and wherein the first initial spatial filter matrix is equal to the first spatial filter matrix from which the first steering matrix is rejected.

38. The apparatus of claim 36, further comprising:

a second initial spatial filter matrix determination module that determines a second initial spatial filter matrix for a third transmission span based on the second spatial filter matrix; and

a third spatial filter matrix derivation module that derives a third spatial filter matrix for the third transmission span based on the second initial spatial filter matrix.