Space-time coding using estimated channel information

Abstract

The invention is directed to techniques for space-time coding in a wireless communication system in which the transmitter makes use of multiple transmit antennas. The transmitter uses channel information estimated by a receiving device and returned to the transmitter, e.g., as feedback. In one exemplary embodiment, the transmitter receives a mean feedback information that defines a mean channel value associated with the different channels of the different antennas. In another exemplary embodiment, the transmitter receives covariance feedback, e.g., statistical values associated with each of the different channels.

Description

TECHNICAL FIELD

[0003] The invention relates to wireless communication and, more particularly, to coding techniques for multi-antenna transmitters.

BACKGROUND

[0004] Space-time coding using multiple transmit-antennas has been recognized as an attractive way of achieving high data rate transmissions with diversity and coding gains in wireless applications. For example, multi-antenna transmitters can offer significant diversity and coding advantages over single antenna transmitters. A number of space-time coding transmitter designs have been developed.

[0005] Most conventional space-time coding transmitters are designed for the scenario where the propagation channels are deterministically known. In practical wireless systems, however, propagation channels are typically not known at the transmitter. Moreover, in practical wireless systems, propagation channels can change over time, with changes in settings of the wireless devices or movement of one wireless device relative to the other wireless device, e.g., movement of a mobile unit relative to a base station.

SUMMARY

[0006] In general, the invention is directed to space-time coding techniques for wireless communication systems in which the transmitter makes use of multiple transmit antennas. As described in greater detail below, the transmitter uses channel information estimated by a receiving device and returned to the transmitter, e.g., as feedback. In other words, the channel information is estimated at the receiver and returned to the transmitter for use in subsequent transmissions to that the signals can be coded in an improved manner.

[0007] In one exemplary embodiment, the transmitter makes use of a mean feedback information that defines a mean channel value associated with the different channels of the different antennas or different multi-paths from one or more antennas. In another exemplary embodiment, the transmitter makes use of covariance feedback, e.g., statistical values associated with the different channels. The mean feedback may be particularly useful when the channels are slow time-varying channels, and the covariance feedback may be particularly useful when the channels are rapid time-varying channels. In other words, if the channels change slowly, the mean feedback can be very useful, but if the channels change rapidly the covariance feedback may be more useful.

[0008] In one embodiment, the invention provides a method comprising receiving estimated channel information for a space-time wireless communication system, and coding signals for transmission by a multiple antenna transmitter based on the estimated channel information.

[0009] In another embodiment, the invention provides a wireless device comprising a coding unit to select symbols based on received channel information estimated for a space-time wireless communication system, and multiple transmit antennas to send the symbols.

[0010] In some embodiments, the invention can be implemented in software. In that case, the invention may be directed to a computer readable medium comprising computer readable instructions that when executed in a wireless device cause the device to code signals for transmission by a multiple antenna transmitter in a space-time wireless communication system based on received channel information estimated by a receiving device.

[0011] In another embodiment, the invention provides a wireless device comprising means for receiving estimated channel information for a space-time wireless communication system, and means for coding signals for transmission by a multiple antenna transmitter based on the estimated channel information.

[0012] In another embodiment, the invention provides a space-time wireless communication system comprising a first wireless device that estimates channel information based on a received signal and transmits the channel information, and a second wireless device that receives the estimated channel information from the first wireless device and codes signals for subsequent transmission via multiple transmit antennas based on the estimated channel information.

[0013] In another embodiment, the invention provides a method comprising receiving communications from a transmitting device via multiple communication channels associated with multiple transmit antennas of the transmitting device, computing estimated channel information for the multiple channels, and communicating the estimated channel information to the transmitting device to control coding of signals for transmission by the multiple antennas.

[0014] In another embodiment, the invention provides a wireless device comprising means for estimating channel information for a space-time wireless communication system, and means for communicating the estimated channel information to a transmitter for use in transmitting subsequent signals by multiple antennas.

[0015] In another embodiment, the invention provides a method comprising receiving estimated channel information associated with multiple channels of a wireless communication signal, and coding subsequent signals for transmission based on the estimated channel information.

[0016] The invention may be capable of providing certain advantages. Specifically, the invention can improve the performance of wireless communication. Numerous embodiments and mathematical techniques are outlined in greater detail below, which can achieve varying levels of performance. In some cases, trade-offs between performance and complexity can be made to meet a specific level of performance and a specific level of complexity.

[0017] The details of one or more embodiments of the invention are set forth in the accompanying drawings and the description below. Other features, objects, and advantages of the invention will be apparent from the description and drawings, and from the claims.

BRIEF DESCRIPTION OF DRAWINGS

[0018] FIG. 1 is a block diagram of a space-time wireless communication system according to an embodiment of the invention.

[0019] FIG. 2 is another block diagram of a space-time wireless communication system according to an embodiment of the invention.

[0020] FIG. 3 is a block diagram illustrating multiple antennas of a transmitter in accordance with an embodiment of the invention.

[0021] FIG. 4 is another block diagram of a space-time wireless communication system according to an embodiment of the invention.

[0022] FIG. 5 is a block diagram of a transmitting device which includes a space-time block coding unit, a set of power loaders, a beam-forming unit, and a set of antennas in accordance with an embodiment of the invention.

[0023] FIGS. 6-16 are graphs illustrating results of simulations of various embodiments of the invention.

DETAILED DESCRIPTION

[0024] The invention is directed to transmitter designs for space-time coding in which the transmitter makes use of multiple transmit antennas. The transmitter uses channel information estimated by a receiving device and returned to the transmitter, e.g., as feedback. In some embodiments outlined in greater detail below, the transmitter makes use of mean feedback information that defines a mean channel value associated with the channels of the different antennas or different multi-paths from one or more antennas. In other embodiments outlined in greater detail below, the transmitter makes use of covariance feedback, e.g., statistical values associated with the different channels. The mean feedback may be particularly useful when the channels are slow time-varying channels, and the covariance feedback may be particularly useful when the channels are rapid time-varying channels. In other words, if the channels change slowly the mean feedback can be very useful, but if the channels change rapidly the covariance feedback may be more useful.

[0025] FIG. 1 is a simplified block diagram of a space-time wireless communication system 10 including a transmitting device 12 (also referred to as transmitter 12) and a receiving device 14 (also referred to as receiver 14). In accordance with space time coding, transmitting device 12 codes signals and transmits the signals via multiple antennas 15A, 15B, 15C. Receiving device 14 includes antenna 16 for receiving signals from device 12. In some cases, receiving device 14 may also include multiple antennas, but the invention is not limited in that respect.

[0026] Transmitting device 12 and receiving device 14 may comprise any of a wide variety of wireless devices that communicate with one another. For example, one of devices 12, 14 may comprise a mobile device and the other of devices 12, 14 may comprise a base station, e.g., in a digital cellular communication system. Alternatively, one of devices 12, 14 may comprise a wireless computer and the other may comprise a wireless network access point, e.g., in a wireless networking setting. In addition, in other applications, each of devices 12, 14 may comprise direct two-way communication devices. In general, system 10 may comprise any of a wide variety of wireless communication systems which could benefit from the feedback techniques described herein.

[0027] In accordance with the invention, receiving device 14 measures channel information, such as the fading amplitudes of the various channels associated with transmission antennas 15A, 15B, 15C. Receiving device 14 sends this measured channel information back to transmitting device 12 so that subsequent signals can be coded based on the measured channel information. In other words, the invention provides a feedback technique in which channel information for multiple space-time channels collected at receiving device 14 is returned to transmitting device 12 for use in subsequent transmissions. In some examples, the channel information includes a mean channel value of the channels associated with the different transmit antennas 15A, 15B, 15C. In other examples covariance feedback is used in which the channel information includes statistical values associated with the different channels.

[0028] The signals transmitted between devices 12, 14 may comprise single carrier signals, or multi-carrier signals. Any of a wide variety of modulation techniques can be used, including, for example, code division multiple access (CDMA), time division multiple access (TDMA), frequency division multiple access (FDMA), orthogonal frequency division multiplexing (OFDM), various other modulation techniques, or even two or more modulation techniques.

[0029] In a space-time wireless system (such as system 10) with Nt transmit antennas and Nr receive antennas, the antenna coefficients can be collected into channel matrix H, with the (μ, ν)th entry as hμν. For each receive antenna ν, the vector: hν:=[h1ν, . . . , hNtν]T can be defined. The columns of H can be concatenated into one channel vector as: 1$\begin{array}{cc}h=\mathrm{vec}\left(H\right)=\left[\begin{array}{c}{h}_1\\ ⋮\\ h_{N_r}\end{array}\right]& \left(1.1\right)\end{array}$

[0030] With perfect channel state information, transmitter 12 knows each realization of h. However, with partial channel state information (CSI), transmitter 12 has some uncertainties on the channel realization h. The uncertainties can be modeled as unknown perturbations around the nominal channel. Specifically, conditioned on channel feedback, transmitter 12 perceives a “nominal-plus-perturbation” channel model as:

{haeck over (h)}={overscore (h)} +ε,   (1.2)

[0031] where {overscore (h)} is deterministic known per feedback realization, and ε is a random vector capturing all uncertainties about h. The partial channel knowledge about the channel will then include the nominal channel {overscore (h)}, and the statistical description on the perturbation error ε. We here use {haeck over (h)} to differentiate the channel perceived at the transmitter from the true channel h; however they have quite different statistical properties. The perception at the transmitter will be updated every time new feedback information becomes available.

[0032] The matrix corresponding to EQUATION 1.2 is:

{haeck over (H)}={overscore (H)} +Ξ,   (1.3)

[0033] where {overscore (H)} and Ξ contain the nominal values and unknown perturbations to describe the Nt×Nr channel matrix H.

[0034] In practice, the perturbation errors may not be Gaussian distributed. But a Gaussian assumption will greatly simplify the transmitter design. The resulting closed-form solutions provide much insight on transmitter optimization based on partial channel knowledge. Hence, for convenience and simplicity, we model {haeck over (h)} as a Gaussian random vector. The statistical property is then described by the mean and covariance matrix of {haeck over (h)}. Specifically, based on channel feedback, the transmitter perceives a random channel distribution as:

{haeck over (h)}˜CN ({overscore (h)}, Σh).   (1.4)

[0035] EQUATION 1.4 provides the general model with Gaussian assumption on the uncertain perturbation errors. We next specify two simplified models, termed as mean feedback and covariance feedback, respectively. In mean feedback, all entries of ε are assumed to be independent from each other, but having the same covariance σε2. Specifically,

{haeck over (h)}˜CN ({overscore (h)}, σε2INtNr).   (1.5)

[0036] Channel mean feedback is suitable to slowly time-varying channels, where instantaneous channel values are fed back. The same uncertainty on all channel coefficients is assumed for simplicity.

[0037] We next highlight several possibilities where channel mean feedback can be realized in practice. We illustrate how to obtain ({overscore (h)}, σε2) based on feedback information.

[0038] Case 1 (Ricean fading channels): In this case, there exist a line-of-sight (LOS) path and diffusing non-LOS paths between transmitter 12 and receiver 14. Hence, the true channel itself is Ricean distributed. We further assume that the diffusing components of all channel coefficients are uncorrelated but with identical variance σh2. Hence,

h˜CN (μh, σh2INtNr),   (1.6)

[0039] where μh contains the channel coefficients corresponding to the LOS paths. In this scenario, we assume that receiver 12 feeds back to transmitter 14 the instantaneous values for the LOS paths and the variance of the diffusing components, without errors. We thus have

{overscore (h)}=μh, σε2=σh2,   (1.7A and 1.7B)

[0040] in the channel mean feedback model.

[0041] Case 2 (delayed feedback): Here we assume that: i) the channel coefficients are slowly time varying according to Jakes' model with Doppler frequency ƒd: ii) antennas 15 are well separated. The channel coefficients are i.i.d. Gaussian distributed as h˜CN (0, σh2INt); and, iii) the channel is acquired perfectly at receiver 14 and is fed back to transmitter 12 via a noiseless channel with delay τ. Let ĥƒ denote the channel feedback. Notice that both h and ĥƒ are complex Gaussian vectors, drawn from the same distribution CN (0, σh2INt).

[0042] It can be shown that E{hĥƒH}=ρσh2INt, where the correlation coefficient ρ:=J0(2πƒdτ) determines the feedback quality. The minimum mean-square error (MMSE) estimator of h based on ĥƒ is given by E{h|ĥƒ}=ρĥƒ, with estimation error having covariance matrix σh2(1−|ρ|2)INtNr. Thus, for each realization of ĥƒ=ĥƒ, 0, the transmitter obtains:

{overscore (h)}=ρĥƒ, 0, σε2=σh2(1−|ρ|2).   (1.8A and 1.8B)

[0043] The deterministic values of {overscore (h)} are updated when the next feedback becomes available.

[0044] Case 3 (quantized feedback): In this case, we assume that the channel is acquired at receiver 14, and is quantized to 2b code words {a(j)}j=12b. The quantizer output is then encoded by b information bits, which are fed back to transmitter 12 with a negligible delay over a noiseless low-speed feedback channel. We assume that transmitter 12 has the same code book, and reconstructs the channel as a(j), if the index j is suggested by the received b bits. Although the quantization error is non-Gaussian and non-white in general, we assume that the quantization errors can be approximated by zero-mean and white Gaussian noise samples, in order to simplify the transmitter design. With εQ2 denoting the approximate variance of the quantization error, the parameters in (1.5) are:

{overscore (h)}=a(j), if index j is received, σε2=εQ2.   (1.9A and 1.9B)

[0045] In addition to Cases 1-3, channel prediction based on pilots inserted at transmitter 12 is also another realization of the general notion of “mean feedback”. Notice that channel predictors take both the feedback delay and the estimation errors into account.

[0046] In covariance feedback, we assume that the channel h varies too rapidly for transmitter 12 to track its instantaneous value. In this case, the channel mean is set to zero, and the relative geometry of the propagation paths manifests itself in a nonwhite covariance matrix Σh. Specifically, we simplify (1.4) to

{haeck over (h)}˜CN (0NtNr×1, Σh).   (1.10)

[0047] The statistical information Σh needs to be updated infrequently.

[0048] Through field measurements, ray-tracing simulations, or using physical channel models, transmitter 12 can acquire such statistical CSI a priori. For certain applications such as fixed wireless, the spatial fading correlations can be determined from such physical parameters as antenna spacing, antenna arrangement, angle of arrival, and angle spread. Likewise, for systems employing polarization diversity, second-order channel statistics will involve the correlation between differently polarized transmissions. Alternatively, receiver 14 can estimate the channel correlations by long-term averaging of the channel realizations, and feed them back reliably to transmitter 12 through a low data rate feedback channel. In applications involving Time Division Duplex (TDD) protocols, transmitter 14 can also obtain channel statistics directly since the forward and backward channels share the same physical (and statistically invariant) channel characteristics even when the time separation between the forward and the backward link is long enough to render the deterministic instantaneous channel estimates outdated. In Frequency Division Duplex (FDD) systems with small angle spread, the downlink channel covariance estimates can be also obtained accurately from the uplink channel covariance through proper frequency calibration processing.

[0049] EQUATION 1.10 specifies a general correlation model for Rayleigh fading channels. However other simplifications can be implemented based on particular propagation environments. EQUATION 1.10 can be further simplified by considering an application scenario where the base station (BS) is unobstructed, and the subscriber unit (SU) is surrounded by rich local scatterers. In this case, the receive antennas are uncorrelated, and the transmit correlation for each receive antenna ν are identical with

Σ0=E{hνhνH}, ∀ν,   (1.11)

[0050] Again, in FIG. 1, either of transmitting device 12 or receiving device 14 can comprise the base station or the mobile unit.

[0051] Σ0 is an arbitrary Hermitian matrix. It turns out, that the antenna spacing at the SU is much smaller (one or two orders of magnitude) than that at the BS, to yield uncorrelated channels among different antennas. In this simplified scenario, we have

Σh=INr{circle over (×)}Σ0.   (1.12)

[0052] Albeit restrictive for the case with multiple receive antennas, the considered model in (1.12) is the most general for the single receive-antenna case, with Σh=Σ0.

[0053] FIG. 2 illustrates a relatively simple design of transmitter 20, which may correspond to transmitter device 12 (FIG. 1). Also depicted in FIG. 2 is a receiver 24. Transmitter 20 includes a set of preorders 21A-21B and a set of antennas 22A-22B.

[0054] Transmitter 20 spreads the information symbol over both space and time. On each antenna (22) μ, a length P spreading code cμ:=[cμ(0), . . . , cμ(P−1]T is used. Different antennas 22 use different spreading codes. This is inherently a low rate system, since only one information symbol is transmitted in P time slots. We will first look at this simple system.

[0055] Define the P×Nt space time matrix C:=[c1, . . . , cNt], and collect the received samples corresponding to each information symbol into a P×Nr matrix Y. The channel input-output relationship is

Y=sCH+W 1   (1.13)

[0056] where W contains additive white Gaussian noise (AWGN) with each entry having variance N0.

[0057] The receiver weights the contribution from each entry of Y to form a decision variable as:

ŝ=tr{GHY}  (1.14)

[0058] Based on (1.13), a desirable maximum ratio combiner (MRC) can be found as:

Gopt=CH   (1.15)

[0059] The signal to noise ratio (SNR) at the MRC output is 2$\begin{array}{cc}\gamma =\frac{E\left\{{\&LeftBracketingBar;\mathrm{tr}\left(H^H{C}^H\mathrm{CH}\right)s\&RightBracketingBar;}^2\right\}}{E\left\{{\&LeftBracketingBar;\mathrm{tr}\left(H^H{C}^{H}W\right)\&RightBracketingBar;}^2\right\}}=\mathrm{tr}\left(H^H{C}^H\mathrm{CH}\right)\frac{E_s}{N_0},& \left(1.16\right)\end{array}$

[0060] where Es:=E{|s|2} is the average energy of the underlying signal constellation.

[0061] For each realization of γ, the instantaneous symbol error rate (SER) is 3$\begin{array}{cc}{P}_{s,\mathrm{PSK}}\left(\gamma \right)=\frac{1}{\pi }{\int }_0^{\frac{\left(M-1\right)\pi }{M}}\exp \left(-\gamma \frac{g_{\mathrm{PSK}}}{{\sin }^2\theta }\right)d\theta ,& \left(1.17\right)\\ P_{s,\mathrm{QAM}}\left(\gamma \right)=\frac{b_{\mathrm{QAM}}}{\sqrt{M}}{\int }_0^{\frac{\pi }{4}}\exp \left(-\gamma \frac{g_{\mathrm{QAM}}}{{\sin }^2\theta }\right)d\theta +b_{\mathrm{QAM}}{\int }_{\frac{\pi }{4}}^{\frac{\pi }{2}}\exp \left(-\gamma \frac{g_{\mathrm{QAM}}}{{\sin }^2\theta }\right)d\theta ,& \left(1.18\right)\end{array}$

[0062] where bQAM:=4(1−1/{square root}{square root over (M)})/π, and the constellation-specific constant g is defined as: 4$\begin{array}{cc}{g}_{\mathrm{PSK}}={\sin }^2\left(\frac{\pi }{M}\right)\mathrm{for}M\text{-}\mathrm{ary}\mathrm{PSK},& \left(1.19\right)\\ g_{\mathrm{QAM}}=\frac{3}{2\left(M-1\right)}\mathrm{for}M\text{-}\mathrm{ary}\mathrm{QAM}.& \left(1.20\right)\end{array}$

[0063] Hence, if the channel is perfect known at transmitter 20, the SER performance can be easily determined from (1.17) and (1.18).

[0064] However, transmitter 20 has only partial knowledge {haeck over (h)}. Transmitter 20 views that the real channel h will be just one realization of {haeck over (h)} during this feedback interval. Since {haeck over (h)}, and thus h, is random, we thus need to average the SER in (1.17) and (1.18) over all possible γ.

[0065] For notation simplicity, define the matrix 5$\begin{array}{cc}Z:=\left(I\otimes C^{H}C\right)\frac{E_s}{N_0}.& \left(1.21\right)\end{array}$

[0066] Using the identity

tr (ABC)=vecH(AH(I{circle over (×)}B)vec(C),   (1.22)

[0067] we simplify the SNR in (1.16) to

γ=hHZh.   (1.23)

[0068] We need to use the following identity: 6$\begin{array}{cc}{E}_z\left\{\exp \left(-z^{H}Az\right)\right\}=\frac{\exp \left(-{\mu }^H{A\left(I+\sum A\right)}^{-1}\mu \right)}{\&LeftBracketingBar;I+\sum A\&RightBracketingBar;},& \left(1.24\right)\end{array}$

[0069] where A is an arbitrary matrix, and z˜CN (μ, Σ). Averaging over γ based on partial CSI, we obtain 7$\begin{array}{cc}{P}_{s,\mathrm{PSK}}=\frac{1}{\pi }{\int }_0^{\frac{\left(M-1\right)\pi }{M}}\frac{\exp \left(-h^{-H}{g^Z\left[I{\sin }^2\theta +g{\sum }_{h}Z\right]}^{-1}\stackrel{\_}{h}\right)}{\&LeftBracketingBar;I{\sin }^2\theta +g{\sum }_{h}Z\&RightBracketingBar;}d\theta & \left(1.25\right)\\ P_{s,\mathrm{QAM}}=\frac{b_{\mathrm{QAM}}}{\sqrt{M}}{\int }_0^{\frac{\pi }{4}}\frac{\exp \left(-h^{-H}{g^Z\left[I{\sin }^2\theta +g{\sum }_{h}Z\right]}^{-1}\stackrel{\_}{h}\right)}{\&LeftBracketingBar;I{\sin }^2\theta +g{\sum }_{h}Z\&RightBracketingBar;}d\theta +b_{\mathrm{QAM}}{\int }_{\frac{\pi }{4}}^{\frac{\pi }{2}}\frac{\exp \left(-h^{-H}{g^Z\left[I{\sin }^2\theta +g{\sum }_{h}Z\right]}^{-1}\stackrel{\_}{h}\right)}{\&LeftBracketingBar;I{\sin }^2\theta +g{\sum }_{h}Z\&RightBracketingBar;}d\theta ,& \left(1.26\right)\end{array}$

[0070] where g takes values in (1.19) and (1.20).

[0071] Once the matrix C is given, transmitter 20 can evaluate the average performance based on on partial CSI in (1.4). We will use this to evaluate the exact SER performance in our numerical results. However, these exact SER expressions are not convenient for transmitter design. We thus rely on an upper bound on the SER.

[0072] By observing that the integrand in (1.25) and (1.26) peaks at θ=π/2, we obtain a unifying SER upper bound as:

P s, bound =αexp (−{overscore (h)}HgZ|I+gΣhZ|−1{overscore (h)})|I+gΣhZ|−1,   (1.27)

[0073] where α: 8$\alpha :=\frac{M-1}{M}.$

[0074] Without any constraint, maximizing Ps, bound leads to the trivial solution that requires infinite power to be transmitted. A practical constraint that takes into account limited budget resources is the average transmitted power, which is expressed as P0=E{tr{(sC)H(sC)}}=Estr{CHC}. Without loss of generality, we assume that P0=Es, and tr{CHC)}=1; i.e., the total transmitted power is Es per symbol. An optimization problem can be formulated as:

Copt=argmin Ps, bound.

C; tr{CHC}=1   (1.28)

[0075] For any matrix, we perform the singular value decomposition (SVD) to obtain 9$\begin{array}{cc}C={\mathrm{\Phi \Delta }}^{\frac{1}{2}}{U}_c^H,& \left(1.29\right)\end{array}$

[0076] where Δ:=diag(δ1, . . . , δNt). Notice that as long as P≧Nt and ΦH Φ=INt, we have

CHC=UcΔUcH,   (1.30)

[0077] thus, the choice of Φ does not affect the performance. We first assume that this is indeed the case, and look for optimal Uc and Dc.

[0078] With Σh=σε2INtNr in mean feedback, we deduce from (1.21) and (1.30): 10$\begin{array}{cc}{\mathrm{gZ}\left(I+g{\sum }_{h}Z\right)}^{-1}=I_{N_r}\otimes \left[\frac{\beta }{{\sigma }_c^2}{U}_c\Delta {U_c\left(I+\beta U_c\Delta U_c^H\right)}^{-1}\right],& \left(1.31\right)\end{array}$

[0079] where the constant β is defined as: 11$\begin{array}{cc}\beta =g{\sigma }_{\epsilon }^2\frac{E_{\delta }}{N_0},& \left(1.32\right)\end{array}$

[0080] We thus simplify the bound in (1.27) to: 12$\begin{array}{cc}{P}_{8,\mathrm{bound}}=\frac{\alpha }{{\&LeftBracketingBar;I_{N_t}+\beta D_c\&RightBracketingBar;}^{N_r}}\exp \left(\frac{1}{{\sigma }_{\epsilon }^2}tr\left\{U_c^H{\stackrel{\_}{\mathrm{HH}}}^H{U}_c\beta {D_c\left(I_{N_t}+\beta D_c\right)}^{-1}\right\}\right).& \left(1.33\right)\end{array}$

[0081] The simplified SER bound in (1.33) allows us to find optimal Uc and Δ in closed form, as detailed in the following.

[0082] We first decompose {overscore (H)}{overscore (H)}H as:

{overscore (HH)}H=UHAUHH, A:=diag(λ1, λ2, . . . , λNt),   (1.34)

[0083] where, without loss of generality, the eigenvalues are arranged in a non-increasing order: λ1≧λ2≧ . . . ≧λNt.

[0084] Substituting (1.33) into (1.28), we can establish that a desirable Uc is:

Uc, opt=UH.   (1.35)

[0085] With the Uc in (1.35), we simplify (1.33) to: 13$\begin{array}{cc}{P}_{8,\mathrm{bound}}={\alpha \left[\prod _{\mu =1}^{N_t}\frac{1}{1+{\delta }_{\mu }\beta }\exp \left(\frac{-{\kappa }_{\mu }{\delta }_{\mu }\beta }{1+{\delta }_u\beta }\right)\right]}^{N_r}.& \left(1.36\right)\\ \mathrm{where}{\kappa }_{\mu }=\frac{{\lambda }_{\mu }}{N_r{\sigma }_{\epsilon }^2}.& \left(1.37\right)\end{array}$

[0086] Δ can be found in closed-form when Nr=1, where we have only one non-zero eigen-value λ1. Since ln(.) is a monotonically increasing function, we define 14$\begin{array}{cc}{\epsilon }_1:=\ln P_{8,\mathrm{bound}}=\ln \alpha -\sum _{\mu =1}^{N_t}\ln \left(1+{\delta }_{\mu }\beta \right)-\frac{{\lambda }_1{\delta }_1\beta }{{\sigma }_{\epsilon }^2\left(1+{\delta }_1\beta \right)}.& \left(1.38\right)\end{array}$

[0087] Our equivalent constrained optimization problem is:

Δopt=argmin ε1

Δ≧0; tr(Δ)=1   (1.39)

[0088] We adopt the special notation |x|+:=max (x, 0). We define several constants as: 15$\begin{array}{cc}\begin{array}{c}a:={\left(1+\frac{N_t}{\beta }\right)}^2,c:=N_t\left(N_t-1\right),\\ b:=\left[\frac{\lambda }{{\mathrm{\beta \sigma }}_{\epsilon }^2}+\left(1+\frac{N_t}{\beta }\right)\left(2N_t-1\right)\right].\end{array}& \left(1.40\right)\end{array}$

[0089] The final solution can then be expressed as: 16$\begin{array}{cc}\begin{array}{c}{\delta }_2=\dots ={\delta }_{N_t}={\left[\frac{2a}{b+\sqrt{b^2-4ac}}-\frac{1}{\beta }\right]}_{+}\\ {\delta }_1=1-\left(N_t-1\right){\delta }_2.\end{array}& \left(1.41\right)\end{array}$

[0090] We now derive an approximate solution for Δ, that is applicable to any Nr. First, we recognize that 17$\begin{array}{cc}E\left\{\exp \left(-\gamma \right)\right\}=\frac{1}{1+{\delta }_{\mu }\beta }\exp \left(-\frac{{\kappa }_{\mu }{\delta }_{\mu }\beta }{1+{\delta }_{\mu }\beta }\right),& \left(1.42\right)\end{array}$

[0091] where {square root}{square root over (γ)} is Rician distributed with Rician factor Kμ and power (1+Kμ)δμβ. It is well known that one can approximate well a Ricean distribution with Ricean factor Kμ by a Nakagami-m distribution with mμ to be: 18$\begin{array}{cc}{m}_{\mu }=\frac{{\left(1+{\kappa }_{\mu }\right)}^2}{1+2{\kappa }_{\mu }}.& \left(1.43\right)\end{array}$

[0092] Note that a Ricean distribution with Kμ=0 coincides with a Nakagami distribution having mμ=1, and both reduce to a Rayleigh distribution. For a Nakagami random variable {square root}{square root over (γ′)} with power (1+Kμ)δμβ, we have 19$\begin{array}{cc}E\left\{\exp \left(-{\gamma }^{\prime }\right)\right\}={\left(1+\frac{\left(1+{\kappa }_{\mu }\right){\delta }_{\mu }\beta }{m_{\mu }}\right)}^{-m_{\mu }}.& \left(1.44\right)\end{array}$

[0093] Approximating {square root}{square root over (γ)} with Ricean distribution by {square root}{square root over (γ′)} with Nakagami distribution, we obtain: 20$\begin{array}{cc}{P}_{s,\mathrm{bound}}\approx {\tilde{P}}_{s,\mathrm{bound}}={\alpha \left[\prod _{\mu =1}^{N_t}{\left(1+\frac{\left(1+K_{\mu }\right){\delta }_{\mu }\beta }{m_{\mu }}\right)}^{-m_{\mu }}\right]}^{N_r}.& \left(1.45\right)\end{array}$

[0094] We define the objective function: 21$\begin{array}{cc}{\epsilon }_2:=\ln {\tilde{P}}_{s,\mathrm{bound}}=\ln \alpha -N_r\left[\prod _{\mu =1}^{N_t}{m}_{\mu }\ln \left(1+\frac{{\delta }_{\mu }\left(1+K_{\mu }\right)\beta }{m_{\mu }}\right)\right].& \left(1.46\right)\end{array}$

[0095] Our optimization problem is approximated by:

Δopt=arg min ε2

Δ≧0; tr(Δ)=1   (1.47)

[0096] Solving (1.46) using the Lagrange method, we have 22$\begin{array}{cc}{\delta }_{\mu }={\left[\frac{m_{\mu }}{Ϛ}-\frac{m_{\mu }}{\left(1+{\kappa }_{\mu }\right)\beta }\right]}_{+}.& \left(1.48\right)\end{array}$

[0097] where ζ is the Lagrangian multiplier, which can be solved by the power constraint tr(Δ)=1.

[0098] Suppose that the final solution ends up with {overscore (N)}t non-zero eigenvalues in Δ. Thus, we have δμ=0, for μ≧{overscore (N)}t+1. For each μ=1, . . . , {overscore (N)}t, we solve ζ using the power constraint to obtain: 23$\begin{array}{cc}{\delta }_{\mu }=\frac{m_{\mu }}{\sum _{l=1}^{{\stackrel{\_}{N}}_t}{m}_l}\left(1+\sum _{l=1}^{{\stackrel{\_}{N}}_t}\frac{m_l}{\left(1+K_l\right)\beta }\right)=\frac{m_{\mu }}{\left(1+K_{\mu }\right)\beta }.& \left(1.49\right)\end{array}$

[0099] To ensure that δ{overscore (N)}t>0, the transmitted power should adhere to the following constraint: 24$\begin{array}{cc}\frac{E_s}{N_0}>\frac{1}{g{\sigma }_e^2}\sum _{l=1}^{{\stackrel{\_}{N}}_t-1}\underset{\underset{:={\gamma }_{\mathrm{th},{\stackrel{\_}{N}}_t}}{⏟}}{\frac{\left({\lambda }_l-{\lambda }_{{\stackrel{\_}{N}}_t}\right)\left(N_r{\sigma }_e^2+{\lambda }_l\right)}{\left(N_r{\sigma }_e^2+{\lambda }_{{\stackrel{\_}{N}}_t}\right)\left(N_r{\sigma }_e^2+2{\lambda }_l\right)}}.& \left(1.50\right)\end{array}$

[0100] From (1.49) and (1.50), we describe the practical algorithm to calculate Δ, in the following steps:

[0101] Step 1: For r=1, . . . , Nt, calculate γth, r from (1.50), based only on the first r eigenvalues of A.

[0102] Step 2: With the given power budget Es ensuring that Es/N0 falls in the interval [γth, r, γth, r+1], set δr+1, . . . , δNt=0, and obtain δ1, . . . , δr according to (1.49) with {overscore (N)}t=r.

[0103] Hence, for any Nr, we obtain a closed-form, albeit approximate, solution. With Nr=1, all the thresholds in (1.50) reduce to: 25$\begin{array}{cc}{\gamma }_{\mathrm{th}}=\frac{{\lambda }_1}{g{\sigma }_e^4}\left(\frac{{\sigma }_e^2+{\lambda }_1}{{\sigma }_e^2+2{\lambda }_1}\right).& \left(1.51\right)\end{array}$

[0104] Hence, the approximate solution reduces to 26$\begin{array}{cc}\begin{array}{c}{\delta }_2=\cdots ={\delta }_{N_t}=\{\begin{array}{cc}{\delta }_2^o& E_s/N_0>{\gamma }_{\mathrm{th}}\\ 0& E_s/N_0\le {\gamma }_{\mathrm{th}}\end{array},\\ {\delta }_1=1-{\delta }_2\left(N_t-1\right),\end{array}& \left(1.52\right)\end{array}$

[0105] where δ20 is simplified from (1.49) as: 27$\begin{array}{cc}{\delta }_2^o=\frac{{\sigma }_e^2\left({\sigma }_e^2+2{\lambda }_1\right)}{N_t{\sigma }_e^2\left({\sigma }_e^2+2{\lambda }_1\right)+{\lambda }_1^2}\left[1+\frac{1}{\beta }\left(N_t-\frac{{\lambda }_1}{{\sigma }_e^2+2{\lambda }_1}\right)\right]-\frac{1}{\beta }.& \left(1.53\right)\end{array}$

[0106] For the case of Nr=1, we will compare the approximate solution in (1.52) with the exact solution in (1.41) later on.

[0107] In the covariance feedback, we only consider the special case of (1.12). Substituting (1.12) into (1.27), we obtain 28$\begin{array}{cc}{P}_{s,\mathrm{bound}}=\alpha {\&LeftBracketingBar;I+{\sum }_0{C}^H\mathrm{Cg}\frac{E_s}{N_0}\&RightBracketingBar;}^{-N_e}.& \left(1.54\right)\end{array}$

[0108] We diagonalize Σ0 using its spectral decomposition:

Σ0=UhDhUhH, Dh:=diag(λ1, . . . , λNt),   (1.55)

[0109] where Uh is unitary, and λμ denotes the μth eigenvalue of Σ0 that is non-negative: λμ≧0. Without loss of generality, we assume that λμ's are arranged in a non-increasing order: λ1≧ . . . ≧λNt≧0. Substituting (1.30) and (1.55) into (1.54), we obtain: 29$\begin{array}{cc}{P}_{s,\mathrm{bound}}=\alpha {\&LeftBracketingBar;I+D_h^{\frac{1}{2}}{U}_h^H{U}_c\Delta U_c^H{U}_h{D}_h^{\frac{1}{2}}g\frac{E_s}{N_0}\&RightBracketingBar;}^{-N_e}& \left(1.56\right)\end{array}$

[0110] For any given Δ, the SER bound in (1.56) will be minimized by the choice of

Uc, opt=Uh,   (1.57)

[0111] based on the Hadamard's inequality.

[0112] With the optimal Uc, we define our objective function as: 30$\begin{array}{cc}\begin{array}{c}{\epsilon }_g:=\ln P_{s,\mathrm{bound}}\\ =-N_r\ln \&LeftBracketingBar;I_{N_2}+D_h\Delta \frac{g{E}_s}{N_0}\&RightBracketingBar;\\ =-N_r\prod _{\mu =1}^{N_t}\ln \left(1+{\lambda }_{\mu }{\delta }_{\mu }\frac{g{E}_s}{N_0}\right).\end{array}& \left(1.58\right)\end{array}$

[0113] In this case, Δ is the solution of

Δopt=arg min ε3.

Δ≧0; tr(Δ)=1   (1.59)

[0114] Solving (1.46) using the Lagrange method, we have 31$\begin{array}{cc}{\delta }_{\mu }={\left[-\frac{1}{Ϛ}-\frac{N_0}{{\lambda }_{\mu }g{E}_8}\right]}_{+}& \left(1.60\right)\end{array}$

[0115] where ζ is the Lagrangian multiplier, which can be solved by the power constraint tr(Δ)=1.

[0116] Suppose that the given power budget, Es supports {overscore (N)}t non-zero δμ's. Solving ζ using the power constraint, which leads to 32$\begin{array}{cc}{\delta }_{\mu }={\left[\frac{1}{\stackrel{\_}{N_t}}+\frac{N_0}{g{E}_8}\left(\frac{1}{\stackrel{\_}{N_t}}\sum _{I=1}^{{\stackrel{\_}{N}}_t}\frac{1}{{\lambda }_I}-\frac{1}{{\lambda }_{\mu }}\right)\right]}_{+}.& \left(1.61\right)\end{array}$

[0117] To ensure δ{overscore (N)}t>0, the transmit power should ensure the following constraint: 33$\begin{array}{cc}\frac{E_s}{N_0}>\frac{1}{g}\underset{\underset{:={\gamma }_{\mathrm{th},\stackrel{\_}{N_t}}}{⏟}}{\left(\frac{\stackrel{\_}{N_t}}{{\lambda }_{\stackrel{\_}{N_t}}}-\sum _{\mu =1}^{{\stackrel{\_}{N}}_t}\frac{1}{{\lambda }_{\mu }}\right)}.& \left(1.62\right)\end{array}$

[0118] Based on (1.62) and (1.61), the practical algorithm for Δ can be summarized in the following steps:

[0119] Step 1: For r=1, . . . , Nt, calculate γth, r from (1.62), based only on the first r channel eigenvalues of Dh.

[0120] Step 2: With the given power budget Es leading to Es/N0 in the interval [γth, r, γth, r+1], set δr+1, . . . , δNt=0, and obtain δ1, . . . , δr according to (1.61) with {overscore (N)}t=r.

[0121] Beam-forming generally refers to a process of multiplying a set of coefficients by the information symbol for transmission on multiple transmit antennas. As depicted in FIG. 3, the information symbols (32) are weighted by the vector u=1, [u . . . , uNt]T before transmission over Nt antennas (33A-33C). The vector u is thus termed a beam. Based on this convention, we can view any precoder C (21A or 21B) in FIG. 2 as a time-varying beamformer. At each time slot p, the pth row of C (denoted as {overscore (c)}p) spreads s into Nt antennas (22A-22B), and thus forming a beam direction along {overscore (c)}p. There are a total of P different beams used for each information symbol, hence redundant time-varying beamforming.

[0122] However, the desirable C based on partial CSI has special structure. First, let us assume Φ=INt. Then, the antenna-steering vector at the pth time slot is

{overscore (c)}p={square root}{square root over (δp)}u*H, p,   (1.63)

[0123] where uH, p is the pth column of UH in mean feedback, or Uh in covariance feedback. Notice that the Nt eigenvectors coincide with the eigenvectors of the channel's correlation matrix perceived at the transmitter,

RH:=E{{haeck over (H)}{haeck over (H)}H}  (1.64)

[0124] For this reason, we term the optimal UH contains eigen beams. Then Δ can be viewed as the power loading matrix onto those eigen beams. Hence, the transmitter adopts the eigen beams in successive time slots, with proper power allocation among them.

[0125] With a general Φ, each row of C is a weighted multiplexing of Nt eigen-beams: 34$\begin{array}{cc}{\stackrel{\_}{c}}_p=\sum _{\mu =t}^{N_t}{\left[\Phi \right]}_{p,\mu }\left(\sqrt{{\delta }_{\mu }}{u}_{H,\mu }^H\right).& \left(1.65\right)\end{array}$

[0126] The power loading on Nt eigen-beams are fixed.

[0127] In summary, the space time spreading as depected in FIG. 2 can be viewed as loaded eigen-beamforming. This explanation is extremely useful, when coupled with space time block coding (STBC).

[0128] FIG. 4 is a block diagram illustrating system 40 which includes a transmitting device (transmitter) 42 and a receiving device (receiver) 44. In particular, transmitting device 42 includes a space-time block coding unit 45 and a beam-forming unit 46 that operate according to the mathematical framework outlined below. Transmitting device 42 of FIG. 4 adopts an orthogonal STBC technique in which multiple symbols are transmitted simultaneously.

[0129] Low-rate is generally inherent to a spread-spectrum scheme. The multiantenna spread spectrum transmitter might be useful for “power-limited” (e.g., military) communication systems, where bandwidth is not at a premium but low transmission power is desired. For “bandwidth-limited” systems on the other hand, it is possible to mitigate the rate loss by sending K>1 symbols, s1, . . . , sK, simultaneously. The rate will then increase to (K/P) symbols/sec/Hz. Notice that our single symbol transmission achieves good performance in an uncoded scenario, which serves as an upper-bound on the performance of multiplexed symbol transmissions. Indeed, when detecting one particular symbol sk, the best scenario happens when all other symbols have been detected correctly, and their effect on sk has been perfectly cancelled.

[0130] One objective is to pursue optimal multiplexing that increases the data rate, without compromising the performance. This may require a symbol separator at the receiver, that does not incur optimality loss. But let us suppose temporarily that such a separator indeed exists, and each symbol is essentially going through separate channels. The desireable precoder Ck for sk will then be

Ck=Φk, Δ1/2UHH; k=1, 2, . . . , K,   (1.66)

[0131] where the Δ is determined as outlined above depending on either channel mean feedback, or, covariance feedback. Because the factor Δ1/2UHH in (1.66) is common ∀k, designing separable precoders is equivalent to selecting separable {Φk}k=1 K matrices. Fortunately, this degree of freedom can be afforded by our design, because so far our Φk's are only required to have orthonormal columns.

[0132] Specifically, we can select Φk from all orthogonal space-time block coding (STBC) matrix. With this choice, our transmitter implements a combination of STBC followed by optimal eigen-beamforming. Here, we focus on complex constellations for brevity; the real constellations can be treated similarly.

[0133] Let skR and skl denote the real and imaginary part of sk, respectively. The following orthogonal STBC designs are available for complex symbols.

[0134] Definition 1: For complex symbols {sk=skR+jskl}k=1K, and P×Nt matrices {Φk, Ψk}k=1K each having entries drawn from {1, 0,−1}, the space time coded matrix 35$\begin{array}{cc}{𝔒}_{N_t}=\sum _{k=1}^K{\Phi }_k{s}_k^R+j\sum _{k=1}^K{\Psi }_k{s}_k^I& \left(1.67\right)\end{array}$

[0135] is termed a generalized complex orthogonal design (GCOD) in variables {sk}k=1K of size P×Nt and rate K/P, if either one of two equivalent conditions holds true:

[0136] i) NtH Nt=(Σk=1K|sk|2)INt, or,

[0137] ii) The matrices {Φk, Ψk}k=1K satisfy the conditions 36$\begin{array}{cc}\begin{array}{ccc}{\Psi }_k^H{\Psi }_k=I_{N_1},& {\Psi }_k^H{\Psi }_k=I_{N_t},& \forall k\\ {\Psi }_k^H{\Psi }_I=-{\Phi }_I^H{\Phi }_k,& {\Psi }_k^H{\Psi }_I=-{\Phi }_I^H{\Phi }_k,& k\ne I\\ {\Psi }_k^H{\Psi }_I={\Phi }_I^H{\Phi }_k,& & \forall k,I\end{array}& \left(1.68\right)\end{array}$

[0138] For complex symbols sk=skR+jskl, we define two precoders corresponding to {Φk, Ψk} as: Ck, 1=ΦkΔ1/2UHH, and Ck, 2=ΨkDc1/2UHH. The combined STBC-Beamforming matrix is now 37$\begin{array}{cc}X=\sum _{k=1}^K{C}_{k,1}{s}_k^R+j\sum _{k=1}^K{C}_{k,2}{s}_k^I={𝔒}_{N_t}{\Delta }^{\frac{1}{2}}{U}_H^H.& \left(1.69\right)\end{array}$

[0139] The received space time matrix is:

Y=XH+W= N t Δ1/2UHHH+W.   (1.70)

[0140] Hence, the original OSTBC matrix Nt now sees an equivalent channel Δ1/2UHHH, as depicted in FIG. 4. By the orthogonal property of OSTBC, each symbol is equivalently passing through a scalar channel of the form

ŝ k =||Δ1/2UHHH||Fsk+wk.   (1.71)

[0141] The SNR is 38$\begin{array}{cc}{\gamma }_k={\&LeftDoubleBracketingBar;{\Delta }^{\frac{1}{2}}{U}_H^{H}H\&RightDoubleBracketingBar;}_F^2\frac{E_s}{N_0}=tr\left\{H^H{U}_H\Delta U_H^{H}H\right\}\frac{E_s}{N_0}.& \left(1.72\right)\end{array}$

[0142] Notice that the SNR from (1.72) is the same as that for the single symbol transmission studied in (1.16), with (1.30), thus, the optimal loading Δ enables space-time block coded transmissions to achieve the performance of single symbol transmission, but with symbol rate K/P. Relative to single symbol transmission, Notice that the OSTBC only have linear complexity increase, relative to the space time spreading transmission described above.

[0143] Utilizing partial channel information at the transmitter, the transmission implement a combination of orthogonal space-time block coding and eigen-beamforming (1.69). Orthogonal space time block coded transmissions are sent using Nt eigen-directions, along the eigenvectors of the correlation matrix of the perceived channels at the transmitter, and are optimally power-loaded. This also justifies our eigen-beamforming interpretation described above.

[0144] For complex symbols, a rate 1 GCOD only exists for Nt=2. It corresponds to the well-known Alamouti code: 39$\begin{array}{cc}{𝔒}_2=\left[\begin{array}{cc}{s}_1& s_2\\ -s_2^{*}& s_1^{*}\end{array}\right]\begin{array}{c}\to \mathrm{space}\\ \downarrow \mathrm{time}\end{array}.& \left(1.73\right)\end{array}$

[0145] For Nt=3, 4, rate 3/4 orthogonal STBC exist, while for Nt>4, only rate 1/2 codes have been constructed. Therefore, for complex symbols, the transmitter of (1.69) achieves good performance with no rate loss only when Nt=2, and pays a rate penalty up to 50%, when Nt>2 and complex constellations are used. To make up for this loss, the transmitter has to enlarge the constellation size, which for the same performance necessitates more transmit-power.

[0146] To tradeoff performance for a constant rate of 1 symbol/sec/Hz, it is possible to send the Alamouti code along the strongest two eigen-beams. Specifically, we construct the 2×Nt space-time coded matrix for the 2D eigen-beamformer: 40$\begin{array}{cc}X=\underset{\underset{{𝔒}_2}{⏟}}{\left[\begin{array}{cc}{s}_1& s_2\\ -s_2^s& s_1^s\end{array}\right]}\underset{\underset{{\Delta }^{1/2}}{⏟}}{\left[\begin{array}{cc}\sqrt{{\delta }_1}& 0\\ 0& \sqrt{{\delta }_2}\end{array}\right]}{\underset{\underset{U_o^H}{⏟}}{\left[u_{H,1},u_{H,2}\right]}}^H.& \left(1.74\right)\end{array}$

[0147] With two eigen beams, the power loading parameters can be calculated with two virtual antennas. To be more specific, we list our answers as follows:

[0148] Case 1—Mean feedback: In 2D beamforming, only power splitting between two basis beams (δ1, δ2) need to be specified. The solution is listed in the following. 41$\begin{array}{cc}{\delta }_1=1-{\delta }_2,\begin{array}{cc}\mathrm{and},& {\delta }_2=\{\begin{array}{cc}{\delta }_2^0& \frac{E_s}{N_0}>{\gamma }_{\mathrm{th},2}\\ 0& \frac{E_s}{N_0}\le {\gamma }_{\mathrm{th},2}\end{array}\end{array}.& \left(1.75\right)\end{array}$

[0149] The threshold is simplified from (1.50) with two virtual antennas: 42$\begin{array}{cc}{\gamma }_{\mathrm{th},2}=\frac{1}{g{\sigma }_{ϵ}^2}\frac{\left({\lambda }_1-{\lambda }_2\right)\left(N_r{\sigma }_{ϵ}^2+{\lambda }_1\right)}{\left(N_r{\sigma }_{ϵ}^2+{\lambda }_2\right)\left(N_r{\sigma }_{ϵ}^2+2{\lambda }_1\right)}.& \left(1.76\right)\end{array}$

[0150] And δ20 is obtained from (1.49) as: 43$\begin{array}{cc}{\delta }_2^0\text{:}=\frac{1+\frac{N_r{\sigma }_{ϵ}^2+{\lambda }_1}{\left(N_r{\sigma }_{ϵ}^2+2{\lambda }_1\right)\beta }+\frac{N_r{\sigma }_{ϵ}^2+{\lambda }_2}{\left(N_r{\sigma }_{ϵ}^2+2{\lambda }_2\right)\beta }}{1+\frac{\left(N_r{\sigma }_{ϵ}^2+2{\lambda }_2\right)}{\left(N_r{\sigma }_{ϵ}^2+2{\lambda }_1\right)}+\frac{\left(N_r{\sigma }_{ϵ}^2+{\lambda }_1\right)}{{\left(N_r{\sigma }_{ϵ}^2+2{\lambda }_2\right)}^2}}-\frac{N_r{\sigma }_{ϵ}^2+{\lambda }_2}{\left(N_r{\sigma }_{ϵ}^2+2{\lambda }_2\right)\beta }& \left(1.77\right)\end{array}$

[0151] The solution in (1.75) reduces to (1.52) if Nt=2 and λ2=0, as expected.

[0152] The threshold is simplified from (1.62) with two virtual antennas: 44$\begin{array}{cc}{\gamma }_{\mathrm{th},2}=\frac{1}{g}\left(\frac{1}{{\lambda }_2}-\frac{1}{{\lambda }_1}\right).& \left(1.79\right)\end{array}$

[0153] The optimal solution is 45$\begin{array}{cc}{\delta }_2^0=\frac{1}{2}\left[1+\frac{N_0}{{\mathrm{gE}}_s}\left(\frac{1}{{\lambda }_1}-\frac{1}{{\lambda }_2}\right)\right].& \left(1.80\right)\end{array}$

[0154] Equations 1.79 and 1.80 are tailored for “Case 2,” in which covariance feedback is used.

[0155] One implementation of this 2D eigen-beam-forming unit is depicted in FIG. 5. In particular, FIG. 5 illustrates a transmitting device 50 which includes a space-time block coding unit 52, a set of power loaders 54A, 54B, a beam-forming unit 56 and a set of antennas 58A, 58B. Specifically, transmitting device 50 operates according to a mathematical framework outlined herein.

[0156] If we set δ2=0 in the 2D beamformer, we reduce (1.74) to: 46$\begin{array}{cc}X=\left[\begin{array}{c}{s}_1{u}_{H·1}^H\\ -s_2^{*}{u}_{H·1}^H\end{array}\right],& \left(1.81\right)\end{array}$

[0157] which corresponds to transmitting information symbols s1 and −s*2 in two consecutive time slots, using the conventional one-dimensional beamforming, along the strongest eigen-vector.

[0158] This leads to following observation: The 2D eigen-beamformer includes the 1D-beamformer as a special case and outperforms it uniformly, without rate reduction, and without essential increase in complexity. Therefore, the 2D eigen-beamformer may be more attractive than the 1D beamformer. It is also worthwhile recalling that the 2D eigen-beamformer is generally better for systems employing Nt=2 transmit-antennas. Because of its full-rate capability and superior performance, the 2D eigen-beamformer may have a major impact in practical systems.

EXAMPLES

[0159] Mean Feedback

[0160] We first consider an uniform linear array with Nt=4 antennas at the transmitter, and a single antenna at the receiver. We consider the delayed channel feedback scenario outlined in Case 2 above, with σh2=1, and a given correlation coefficient ρ. We will present simulation results for two constellations: QPSK (4-PSK), and 16-QAM. Simulation results are averaged over 10,000 Monte-Carlo feedback realizations.

[0161] We first compare optimal power loading based on the Ricean distribution (1.41) with that based on the Nakagami distribution (1.52). FIG. 6 verifies that both approaches have almost identical performance. For this reason, we subsequently plot only the performance of power loading based on (1.52). FIG. 6 also confirms that the SER bound is tight, and has a constant difference with the exact SER across the Es/N0 range considered. This justifies well our approach of pushing down the bound to decrease the SER. across the Es/N0 range considered. This justifies well our approach of pushing down the bound to decrease the SER.

[0162] FIGS. 7 and 8 compare optimal power loading, equal power loading (that has the same performance as plain STBC without beamforming), 1D and 2D beamforming, for both QPSK and 16QAM. When the feedback quality is low (ρ=0.6), FIG. 7 shows that optimal power loading performs close to equal power loading, while it considerably outperforms conventional 1D beamforming. On the other hand, when the feedback quality improves to ρ=0.9, equal power loading is highly suboptimum. The conventional beamforming performs close to the optimal power loading at low SNR, while it becomes inferior at sufficiently high SNR. Notice that the 2D beamformer outperforms the 1D beamformer uniformly. When Es/N0>γth for each feedback realization, although both 2D and 1D beamformer become suboptimal, the 2D beamformer benefits from the order-2 diversity. Since gQPSK/g16QAM=5, we observe that 7.0 dB higher power is required for 16-QAM than QPSK, to adopt Nt directions.

[0163] We next tested our results with multiple receive antennas. FIGS. 9 and 10 are the counterparts of FIGS. 7 and 8, but with Nr=2 receive antennas. It can be seen that the performance of the 2D beamformer coincides with the optimal beamformer for a larger range of Es/N0 than that of the 1D beamformer. This is different from the single receive antenna case, where 2D and 1D beamformers deviate from the optimal beamformer at the same time, since there is only one dominant direction.

[0164] Covariance Feedback

[0165] We consider a uniform linear array with Nt=4 antennas at the transmitter, and a single antenna at the receiver. We assume that the side information including the distance between the transmitter and the receiver, the angle of arrival, and the angle spread are all available at the transmitter. Let λ be the wavelength of a narrowband signal, dt the antenna spacing, and Δ the angle spread. We assume that the angle of arrival is perpendicular to the transmitter antenna array. Thus, using the result of [eq. (6)] for small angle spread, we can simplify the correlation coefficient between the pth and the qth transmit-antenna as follows: 47$\begin{array}{cc}{\left[\sum _0\right]}_{p,q}\approx \frac{1}{2\pi }{\int }_0^{2\pi }\exp \left[-\mathrm{j2}\pi \left(p-q\right)\Delta \frac{d_{\square }}{\lambda }\sin \theta \right]d\theta .& \left(1.82\right)\end{array}$

[0166] Our tests focus on two channels: Channel 1 has dt=0.5λ, and Δ=5°; while Channel 2 has lower spatial correlations with dt=0.5λ, and Δ=25°. Notice that Δ can be calculated from the radius of the local scatterers, and the distance between the transmitter and the receiver [3].

[0167] We present simulations for two constellations: QPSK and 16-QAM. In all the plots, the SNR is defined as the total transmitted power divided by the noise power: SNR=Es/N0.

[0168] FIGS. 11 and 12 show the optimal power allocation among different beams for Channels 1 and 2, for both QPSK and QAM constellations. At low SNR, the transmitter prefers to shut off certain beams, while it approximately equates power to all antennas at sufficiently high SNR to benefit from diversity. Notice that the choice of how many beams are retained depends on the constellation-specific SNR thresholds. For QPSK, we can verify that γth, 2=10.2 dB, and γth, 3=37.5 dB for Channel 1, while γth, 2=−15.0 dB, and γth, 3=8.1 dB for Channel 2. Since gQPSK/g16QAM=5, the threshold γth, r for 16-QAM is 10 log10(5)=7.0 dB higher for QPSK; we observe that 7.0 dB higher power is required for 16-QAM before switching to the same number of beams as for QPSK.

[0169] With Channel 1, FIGS. 13 and 14 depict the exact SER, and the SER upper-bound for: optimal power loading, equal power loading (that has the same performance as plain STBC without beamforming), and 1D beamforming. Since Channel 1 is highly correlated, only r=2 beams are used in the considered SNR range for optimal loading. Therefore, the 2D eigen-beamformer is overall optimal for Channel 1 in the considered SNR range, and its performance curves coincide with those of the optimal loading. FIGS. 13 and 14 confirm that the optimal allocation outperforms both the equal power allocation, and the 1D beamforming. The difference between optimal loading and equal power loading is about 3 dB as SNR increases, since 2 out of 4 beams are so weak that the power allocated to them is wasted. The differences between the upper-bound and the exact SER in FIGS. 13 and 14 justify our approach that pushes down the upper-bound to minimize the exact SER.

[0170] On the other hand, Channel 2 is less correlated than Channel 1, and all four beams are used at high SNR. Equal power loading approaches the optimal loading when SNR is sufficiently high, but is inferior to both the 2D eigen-beamforming and the optimal loading at low to medium SNR, as confirmed by FIGS. 15 and 16. It is also shown that 2D eigen-beamforming outperforms 1D beamforming uniformly, and the difference is quite significant at moderate to high SNR. By checking the eigenvalues of Channel 2, we find that Dh=diag(1.79, 1.74, 0.45, 0.02). Notice that the first two eigenvalues are not disparate enough, and the 1D beamformer is only optimal when Es/N0≦γth, 2=−8.0 dB for 16-QAM. On the other hand, the 2D eigen-beamformer achieves optimality up to Es/N0=γth, 3=15.1 dB for 16-QAM, as seen in FIG. 16. This observation corroborates the importance of 2D eigen-beamforming relative to 1D beamforming.

[0171] Various embodiments of the invention have been described. Nevertheless, various modifications can be made, without departing from the spirit and scope of the invention. For example, other mathematical techniques may be used in the feedback scheme, such as the median value associated with the channels, or the standard deviation associated with the mean channel value. Also, certain aspects of the invention may find use in systems that do not necessarily implement multiple transmit antennas. For example, the techniques have been described above in the context of multiple transmit antennas that define multiple antennas. However, in some cases, a given transmit antenna can define a multi-path signal. In that case, each reception of the multi-path signal can be viewed as a channel. In accordance with the invention, the techniques described herein can also be used to estimate channel information associated with multiple channels of a multi-path signal, and then feed back the estimated channel information for use in generating subsequent signals. In other words, in some embodiments of the invention, a single transmit antenna can be used to create multi-path signals, for which channel information is estimated and feed back to the transmitter for use in generating subsequent signals. Also, in some cases, the transmitter can have multiple transmit antennas, with each antenna sending multi-path signals. In other words, the invention can also apply with multiple transmit antenna channels having a plurality of multi-path channels for each antenna.

[0172] The described techniques can be embodied in a variety of transmitters including base stations, cell phones, laptop computers, handheld computing devices, personal digital assistants (PDA's), and the like. The devices may include a digital signal processor (DSP), field programmable gate array (FPGA), application specific integrated circuit (ASIC) or similar hardware, firmware and/or software for implementing the techniques. In other words, block coding units and Eigen-beam-forming unit, as described herein, may be implemented in such hardware, software, firmware, or the like.

[0173] If implemented in software, a computer readable medium may store computer readable instructions, i.e., program code, that can be executed by a processor or DSP to carry out one of more of the techniques described above. For example, the computer readable medium may comprise random access memory (RAM), read-only memory (ROM), non-volatile random access memory (NVRAM), electrically erasable programmable read-only memory (EEPROM), flash memory, or the like. The computer readable medium may comprise computer readable instructions that when executed in a wireless communication device, cause the wireless communication device to carry out one or more of the techniques described herein. These and other embodiments are within the scope of the following claims.

Claims

1. A method comprising: receiving estimated channel information for a space-time wireless communication system; and coding signals for transmission by a multiple antenna transmitter based on the estimated channel information.

2. The method of claim 1, wherein receiving estimated channel information comprises receiving information defining a mean estimate of multiple channels associated with the multiple antennas.

3. The method of claim 2, further comprising identifying that the channels are substantially slow time-varying channels.

4. The method of claim 1, wherein receiving estimated channel information comprises receiving a perturbation vector defining uncertainties of the channels relative to a nominal vector that nominally defines the channels.

5. The method of claim 1, wherein receiving estimated channel information comprises receiving information defining a statistical estimate of the multiple channels associated with the multiple antennas.

6. The method of claim 5, further comprising identifying that the channels are substantially rapid time-varying channels.

7. The method of claim 1, wherein receiving estimated channel information comprises receiving information defining a mean estimate of multiple channels associated with the multiple antennas and receiving information defining statistical estimates of the multiple channels.

8. The method of claim 7, further comprising: identifying that the channels are substantially slow time-varying channels; and coding signals for transmission by the multiple antenna transmitter using the mean estimate.

9. The method of claim 7, further comprising: identifying that the channels are substantially rapid time-varying channels; and coding signals for transmission by the multiple antenna transmitter using the statistical estimates.

10. The method of claim 1, wherein the method is performed by one of a wireless mobile device and a wireless base station.

11. The method of claim 1, further comprising forming Eigen-beams based on the estimated channel information and transmitting the Eigen-beams via the multiple antennas.

12. A wireless device comprising: a coding unit to select symbols based on received channel information estimated for a space-time wireless communication system; and multiple transmit antennas to send the symbols.

13. The wireless device of claim 12, further comprising a beam-forming unit to form Eigen-beams based on the selected symbols and the received channel information, wherein the multiple transmit antennas send the symbols by sending the Eigen-beams.

14. The wireless device of claim 13, wherein the beam-forming unit comprises a two-dimensional Eigen-beam-forming unit.

15. The wireless device of claim 12, wherein the coding unit applies a block coding matrix:

16. The wireless device of claim 15, wherein the following condition holds true:

17. The wireless device of claim 13, wherein the beam-forming unit forms the Eigen-beams by applying the following:

18. The wireless device of claim 17, wherein:

19. The wireless device of claim 13, wherein the beam-forming unit comprises a two-dimensional Eigen-beam-forming unit.

20. The wireless device of claim 19, wherein the two-dimensional Eigen-beam-forming unit forms Eigen-beams by applying the following:

21. The wireless device of claim 12, wherein the received channel information estimated for the space-time wireless communication system includes a mean estimate of multiple channels associated with the multiple antennas.

22. The wireless device of claim 12, wherein the received channel information estimated for the space-time wireless communication system includes a perturbation vector defining uncertainties of the channels relative to a nominal vector that nominally defines the channels.

23. The wireless device of claim 12, wherein the received channel information estimated for the space-time wireless communication system includes a statistical estimate of the multiple channels associated with the multiple antennas.

24. The wireless device of claim 12, wherein the device comprises one of a wireless mobile device and a wireless base station.

25. A computer readable medium comprising computer readable instructions that when executed in a wireless device cause the device to code signals for transmission by a multiple antenna transmitter in a space-time wireless communication system based on received channel information estimated by a receiving device.

26. The computer readable medium of claim 25, wherein the channel information includes a mean estimate of multiple channels associated with the multiple antennas.

27. The computer readable medium of claim 25, wherein the channel information includes a perturbation vector defining uncertainties of the channels relative to a nominal vector that nominally defines the channels.

28. The computer readable medium of claim 25, wherein the channel information includes a statistical estimate of the multiple channels associated with the multiple antennas.

29. A wireless device comprising: means for receiving estimated channel information for a space-time wireless communication system; and means for coding signals for transmission by a multiple antenna transmitter based on the estimated channel information.

30. The wireless device of claim 29, wherein the estimated channel information includes a mean estimate of multiple channels associated with the multiple antennas.

31. The wireless device of claim 29, wherein the estimated channel information includes a perturbation vector defining uncertainties of the channels relative to a nominal vector that nominally defines the channels.

32. The wireless device of claim 29, wherein the estimated channel information includes a statistical estimate of the multiple channels associated with the multiple antennas.

33. A space-time wireless communication system comprising: a first wireless device that estimates channel information based on a received signal and transmits the channel information; and a second wireless device that receives the estimated channel information from the first wireless device and codes signals for subsequent transmission via multiple transmit antennas based on the estimated channel information.

34. The space-time wireless communication system 33, wherein the estimated channel information includes a mean estimate of multiple channels associated with the multiple antennas.

35. The space-time wireless communication system 33, wherein the estimated channel information includes a perturbation vector defining uncertainties of the channels relative to a nominal vector that nominally defines the channels.

36. The space-time wireless communication system 33, wherein the estimated channel information includes a statistical estimate of the multiple channels associated with the multiple antennas.

37. A method comprising: receiving communications from a transmitting device via multiple communication channels associated with multiple transmit antennas of the transmitting device; computing estimated channel information for the multiple channels; and communicating the estimated channel information to the transmitting device to control coding of signals for transmission by the multiple antennas.

38. The method of claim 37, wherein the estimated channel information includes a mean estimate of multiple channels associated with the multiple antennas.

39. The method of claim 37, wherein the estimated channel information includes a perturbation vector defining uncertainties of the channels relative to a nominal vector that nominally defines the channels.

40. The method of claim 37, wherein the estimated channel information includes a statistical estimate of the multiple channels associated with the multiple antennas.

41. A wireless device comprising: means for estimating channel information for a space-time wireless communication system; and means for communicating the estimated channel information to a transmitter for use in transmitting subsequent signals by multiple antennas.

42. The wireless device of claim 41, wherein the estimated channel information includes a mean estimate of multiple channels associated with the multiple antennas.

43. The wireless device of claim 41, wherein the estimated channel information includes a perturbation vector defining uncertainties of the channels relative to a nominal vector that nominally defines the channels.

44. The wireless device of claim 41, wherein the estimated channel information includes a statistical estimate of the multiple channels associated with the multiple antennas.

45. A method comprising: receiving estimated channel information associated with multiple channels of a wireless communication signal; and coding subsequent signals for transmission based on the estimated channel information.

46. The method of claim 45, wherein receiving estimated channel information comprises receiving information defining a mean estimate of the multiple channels.

47. The method of claim 45, wherein receiving estimated channel information comprises receiving information defining a statistical estimate of the multiple channels.

48. The method of claim 45, wherein the multiple channels include channels associated with multiple transmit antennas.

49. The method of claim 45, wherein the multiple channels include channels of a multi-path signal associated with a transmit antenna.