A multi-input, multi-output (MIMO) link refers to a communication system with multiple antennas at a transmitter and a receiver. MIMO links are a known, effective way to increase link capacity, in terms of bits per second per hertz, of wireless links without increasing the bandwidth and power budget. It has been shown that with n_T transmit antennas on a transmit side and n_R antennas on a receive side, the capacity gain is proportional to n=min [n_T,n_R]. The capacity gain results from the formation of MIMO channels and n spatial channels.
There may be at least two benefits from multiple antenna systems; spatial diversity and beamforming. Spatial diversity refers to the ability to collect independent received samples containing the same piece of transmitted information. By combining different independent observations, the effect of channel fading can be reduced. Beamforming is used to steer the antenna pattern so that a strong beam is oriented toward the desired signal and weak beams or even nulls, are formed for interference signals. It has been shown that when complete feedback of the channel state matrix is available to the transmitter, the optimal transmission scheme for a MIMO link is a combined adaptive power control and eigen-beamforming scheme, based on instantaneous channel fading.
However, complete feedback is not easy to achieve especially when the number of transmit and receive antennas is large.
Exemplary embodiments of the present invention are directed to a method of transmitting or receiving in an at least one-input, at least one-output communication system with constrained feedback information.
In an exemplary embodiment, the present invention is directed to a method of supplying constrained feedback information to a transmitter of an at least one-input, at least one-output communication system, which may include receiving at least one receive signal from at least one transmit antenna at the transmitter, generating the constrained feedback information at a receiver, and transmitting the constrained feedback information back to the transmitter.
In an exemplary embodiment, the present invention is directed to a method of receiving constrained feedback information from at least one receive antenna of a receiver of an at least one-input, at least one-output communication system, which may include transmitting a plurality of transmit signals from at least one transmit antenna of a transmitter, based on a plurality of input signals, a plurality of power gains, and a plurality of beamformer matrices, receiving the constrained feedback information back from the receiver, and selecting at least one of the plurality of power gains and at least one of the plurality of beamformer matrices for a subsequent transmission based on the constrained feedback information.
Exemplary embodiments of the present invention will become more fully understood from the detailed description given below and the accompanying drawings, which are given for purposes of illustration only, and thus do not limit the invention.
It should be emphasized that the drawings of the instant application are not to scale but are merely schematic representations, and thus are not intended to portray the specific dimensions of the invention, which may be determined by skilled artisans through examination of the disclosure herein.
The receiver 300 may include at least one receive antenna 3021 . . . NR, a main decoder 304 (for example, a minimum mean square error decoder (MMSE)), at least one channel decoder 3061 . . . NT, a channel state information receive (CSIR) estimator 308, and a channel state information transmit (CSIT) generator 310.
Elements 210a and 310a illustrate exemplary internal operations of the power and beamforming controller 210 and the CSIT generator 310, respectively, and will be discussed in more detail below.
In operation, each of the encoders 2021 . . . NI may receive a stream of channel symbols Ω1 . . . N, each acting as an independent channel encoder. The output T1 . . . N of each encoder 2021 . . . NI may be input to at least one power control block 2041 . . . NT, which may perform a power adjustment by multiplying the output T1 . . . N by a power gain μ1 . . . NT provided by the power and beamforming controller 210. The power adjusted outputs from the power control blocks 2041 . . . NT may be input to the beamformer 206 along with a beamformer matrix, also supplied by the power and beamforming controller 210. The beamformer 206 may produce transmit signals X1 . . . NT which may be supplied to the transmit antenna 2081 . . . NT.
At the receiver 300, each of the receive antenna 3021 . . . NR may receive a receive signal Y1 . . . NR and forward the receive signal Y1 . . . NR to the main decoder 304 and to the CSIR estimator 308. The CSIR estimator 308 may produce an estimation of the channel matrix V, which may be input to the CSIT generator 310. The CSIT generator 310 may produce a feedback signal U, which may be fed back to the main decoder 304 at the receiver 300 and the power and beamforming controller 210 of the transmitter 200. The decoder 304 may produce a stream of symbols S1 . . . NT which may be input to the channel decoders 3061 . . . NT. The channel decoders 3061 . . . NT may output output streams Ω1 . . . N.
As described above, elements 210a and 310a illustrate exemplary internal operation of the power and beamforming controller 210 and the CSIT generator 310, respectively. In particular, the CSIT generator 310 may receive the channel matrix V from the CSIR estimator 308 and may produce the feedback signal U. As shown in element 310a, the CSIT generator 310 may compare the channel matrix V to one of a plurality of partitions 312a and may select the closest match. Each of the partitions 312a may have a corresponding partition ID 314a, and the partition ID 314a which corresponds to the best match partition 312a may be output as feedback signal U.
In an exemplary embodiment of the present invention, feedback signal U may be constrained feedback information, for example, feedback signal U may be limited to a certain number of bits, such as four bits, two bits, or even one bit. In an exemplary embodiment of the present invention, element 310a may be defined as a feedback strategy.
The power and beamforming controller 310 may receive the feedback signal U from the CSIT generator 310. Element 210a illustrates exemplary internal operation of the power and beamforming controller 210. In particular, the feedback signal U may be received by an index decoder 212a, which may match the feedback signal U, with the power control gain and beamforming matrix 214a which is the closest match. The closest matching power control gain μnT and beamforming matrix bf may then be supplied to the power control blocks 2041 . . . NT and the beamformer 206, respectively. In an exemplary embodiment, element 210a is characterized as a transmission strategy. In exemplary embodiments of the present invention, both the feedback strategy in the transmission strategy may be stored in tables on the receiver side and transmit side, respectively.
In an exemplary embodiment, the present invention may be adaptive from frame to frame, and the feedback signal U may be provided only for a given frame and only for a given partition.
With respect to
In the exemplary embodiment of the
In an exemplary embodiment of the present invention, an iterative algorithm may be used to find the optimal partition and the optimal power control/beam forming matrices. An exemplary iterative algorithm is set forth below.
Initially, a “distance measure” may be defined as:
Further a q-th partition region, Rq, may be defined as the region of channel matrix “nearest” to (ρq,wq)where the distance is defined in (0.1).
Rq={vεZW
A first step of the algorithm may be to obtain Q positive semidefinite matrices {b0, . . . , bQ-1}. This may be accomplished by initializing the ρq and wq by singular value decomposition of bq where ρq is the nT×nT diagonal matrix with diagonal elements corresponding to the nT eigenvalues of bq; wq is the nT×nT hermitian matrix corresponding to the eigenmatrix of bq for all q=[0,Q−1].
As a result, the Q partition regions {R0, . . . , RQ-1} may be initialized according to (0.2).
A second step of the algorithm may be to evaluate the optimal power control matrix and beam forming matrix from the partition region. The beamforming matrix corresponding to the q-th partition, wq, may be the eigenvector matrix of
The power control matrix may be a diagonal matrix given by:
where ψ is the diagonal matrix corresponding to the eigenvalues of
A third step of the algorithm may update the partition region {R0, . . . ,RQ-1} according to (0.2). The second step may be repeated until convergence occurs.
The first through third steps may be repeated for other random initializations {b0, . . . ,bQ-1} and results may be chosen that give the maximum capacity.
Exemplary embodiments of the present invention may be compatible with fading channels with complex Gaussian channel noise. Exemplary embodiments of the present invention may combine adaptive beamforming and power control schemes to improve or optimize the ergodic capacity of a MIMO or MISO link. Exemplary embodiments of the present invention may be directed to a generic framework for the design of an improved adaptive power control and beam-forming scheme with a feedback capacity constraint. As a result of exemplary embodiments of the present invention, the feedback capacity may be better utilized. Also, exemplary embodiments of the present invention may be used to evaluate how far a given heuristic design is from the optimal performance level.
Exemplary embodiments of the present invention may be useful in wireless infrastructures, such as UMTS or CDMA2000, as well as other wireless systems, such as a wireless LAN.
The invention being thus described, it will be obvious that the same may be varied in many ways. Such variations are not to be regarded as departure from the spirit and scope of the exemplary embodiments of the present invention, and all such modifications are intended to be included within the scope of the following claims.
Number | Name | Date | Kind |
---|---|---|---|
6760882 | Gesbert et al. | Jul 2004 | B1 |
6842632 | Raghothaman et al. | Jan 2005 | B2 |
6922445 | Sampath et al. | Jul 2005 | B1 |
7043259 | Trott | May 2006 | B1 |
20020027957 | Paulraj et al. | Mar 2002 | A1 |
20020034263 | Schmidl et al. | Mar 2002 | A1 |
20020187753 | Kim et al. | Dec 2002 | A1 |
20020193146 | Wallace et al. | Dec 2002 | A1 |
20030003863 | Thielecke et al. | Jan 2003 | A1 |
20030220103 | Kim et al. | Nov 2003 | A1 |
20030224750 | Sampath | Dec 2003 | A1 |
20030236080 | Kadous et al. | Dec 2003 | A1 |
20040027994 | Baines | Feb 2004 | A1 |
20060018402 | Mehta et al. | Jan 2006 | A1 |
Number | Date | Country | |
---|---|---|---|
20040242162 A1 | Dec 2004 | US |