The disclosed embodiments relate generally to wireless communication, and, more particularly, to method and user equipment for determining precoder of multi-input multi-output (MIMO) system.
In conventional network of 3rd generation partnership project (3GPP) 5G new radio (NR), the user equipment (UE) can measure channel state information reference signals (CSI-RSs) transmitted from a base station (BS) under a multi-input multi-output (MIMO) network, and determine downlink channel matrices according to the CSI-RSs. Then, the UE can calculate precoder(s) based on the downlink channel matrices, and report the compressed/quantized precoder(s) through precoding matrix indicator(s) (PMIs) to the BS using one of the specified codebooks. Therefore, the BS can transmit subsequent physical downlink shared channels (PDSCHs) by applying the precoder(s). However, the subsequent transmission of PDSCHs applied by the precoder(s) may be serious distorted due to the inaccurate calculation of the precoder(s) calculated and reported by the UE.
Method and user equipment (UE) are provided for determining precoder of multi-input multi-output (MIMO) system. In particular, a base station (BS) can transmit at least one channel state information-reference signal (CSI-RS) to the UE. The UE can receive the at least one CSI-RS, and estimate at least one covariance matrix of at least one downlink channel matrix according to the at least one CSI-RS at different times and frequencies. Then, the UE can transmit the at least one covariance matrix or at least one parameter associated with the least one covariance matrix to the BS for the BS to reconstruct at least one covariance matrix, and determine a precoder according to the at least one reconstructed covariance matrix.
Other embodiments and advantages are described in the detailed description below. This summary does not purport to define the invention. The invention is defined by the claims.
The accompanying drawings, where like numerals indicate like components, illustrate embodiments of the invention.
Reference will now be made in detail to some embodiments of the invention, examples of which are illustrated in the accompanying drawings.
The gNB 121 may provide communication coverage for a geographic coverage area in which communications with the UE 110 is supported via a communication link 101. The communication link 101 shown in the 5G NR network 100 may include UL transmissions from the UE 110 to the gNB 121 (e.g., on the Physical Uplink Control Channel (PUCCH) or Physical Uplink Shared Channel (PUSCH)) or downlink (DL) transmissions from the gNB 121 to the UE 110 (e.g., on the Physical Downlink Control Channel (PDCCH) or Physical Downlink Shared Channel (PDSCH)).
Similarly, for the UE 110, antennas 177 transmit and receives RF signal under MIMO network. RF transceiver module 176, coupled with the antennas, receives RF signals from the antennas, converts them to baseband signals and sends them to processor 173. The RF transceiver 176 also converts received baseband signals from the processor 173, converts them to RF signals, and sends out to antennas 177. Processor 173 processes the received baseband signals and invokes different functional modules and circuits to perform features in the UE 110. Memory 172 stores program instructions and data 170 to control the operations of the UE 110.
Although a specific number of the antennas 177 and 197 are depicted in
The gNB 121 and the UE 110 also include several functional modules and circuits that can be implemented and configured to perform embodiments of the present invention. In the example of
Note that the different functional modules and circuits can be implemented and configured by software, firmware, hardware, and any combination thereof. The function modules and circuits, when executed by the processors 193 and 173 (e.g., via executing program codes 190 and 170), allow the gNB 121 and the UE 110 to perform embodiments of the present invention.
Then, the UE 110 estimates at least one covariance matrix of at least one downlink channel matrix according to the at least one CSI-RS at different times and frequencies. The UE 110 transmits the at least one covariance matrix or at least one parameter associated with the least one covariance matrix to the BS 121.
After receiving the at least one covariance matrix or the at least one parameter associated with the least one covariance matrix, the BS 121 can: (1) derive a precoder according to the at least one covariance matrix; or (2) reconstruct the at least one covariance matrix based on the at least one parameter associated with the least one covariance matrix and derive the precoder according to the at least one covariance matrix.
In some embodiments, the UE 110 estimates the covariance matrices of the downlink channel matrix according to the CSI-RSs at different times and frequencies. In particular, after receiving the CSI-RSs at different times and frequencies, the UE 110 estimates the downlink channel matrices H[n,m] of nR×nT MIMO channels as below:
nR is the number of receiving antennas (i.e., antennas 177). nT is number of transmitting antennas (i.e., antennas 197). n is time domain index. m is frequency domain index.
Then, the UE 110 estimates covariance matrices Ψ[n,m] based on the downlink channel matrices H[n,m] as below:
Ψ[n,m]=HH[n,m]H[n,m]
It is noted that Ψ[n,m] should be ĤH[n,m]Ĥ[n,m] while Ĥ[n,m]=Φ[n,m]H[n,m] and Φ[n,m] is time-frequency offset. Because Φ[n,m] is diagonal matrix with unit-phase factors, ΦH[n,m]Φ[n,m] is I which is an identity matrix. Accordingly,
which means the time-frequency offset Φ[n,m] can be ignored while the UE 110 estimates covariance matrices Ψ[n,m].
In some implementations, as shown in
Then, the BS 121 transmits subsequent PDSCHs to the UE 110 by using the precoders W.
In some implementations, the UE 110 compresses the covariance matrices Ψ[n,m], and transmits at least one parameter associated with the compressed covariance matrices to the BS 121. In particular, ψij[n,m] represents an entry in row i and column j of Ψ[n,m]. Each entry ψij of covariance matrix Ψ[n,m] is considered as a function of time n and frequency m. The entries ψij at different times (N number in total) and frequencies (M number in total) form Ψ[n,m].
The covariance matrices Ψ[n,m] are compressed by: (1) being projected onto two-dimension (2D) time-frequency basis matrices; or (2) being represented as linear combination of 2D sinusoids matrices. A 2D sinusoids matrix has the entry in row n and column m as e−j2πmτej2πnν, where τ models delay and ν models Doppler shift. More specifically, after compression, the compressed covariance matrices can be represented as:
while Bkl is known parameter by both the UE 110 and the BS 121.
The UE 110 selects a subset Sij from the 2D time-frequency basis matrices (or from the 2D sinusoids matrices) and coefficients (cijkl:(k,l)∈Sij) corresponding to the subset, and then, as shown in
Then, the BS 121 reconstructs covariance matrices Ψij by Ψij≈Σ(k,l)∈S
Then, the BS 121 transmits subsequent PDSCHs to the UE 110 by using the precoders W.
In some implementations, the UE 110 compresses the covariance matrices Ψ[n,m], and transmits at least one parameter associated with the compressed covariance matrices to the BS 121. In particular, ψij[n,m] represents an entry in row i and column j of Ψ[n,m]. Each entry ψij of covariance matrix Ψ[n,m] is considered as a function of time n and frequency m. The entries ψij at different frequencies (M number in total) in time n form vectors ψij[n].
The vectors ψij[n] are compressed by: (1) being projected onto one-dimension (1D) frequency basis vectors; or (2) being represented as linear combination of 1D sinusoids vectors. A 1D sinusoids vector has the m-th entry as e−j2πmτ, where τ models delay. More specifically, after compression, the compressed vectors can be represented as:
ψij[n]=Σk=1Mcijk[n]bk
while bk is known parameter by both the UE 110 and the BS 121.
The UE 110 selects a subset Sij[n] from the 1D frequency basis vectors (or from the 1D sinusoids vectors) and coefficients (cijk[n]: k∈Sij[n]) corresponding to the subset Sij[n], and then, as shown in
Then, the BS 121 reconstructs vectors ψij[n] by ψij[n]≈Σk∈S
Then, the BS 121 transmits subsequent PDSCHs to the UE 110 by using the precoders W.
In some implementations, the UE 110 calculates (or approximates) a decomposition of the covariance matrices Ψ[n,m] as:
Ψ[n,m]=UH[n,m]U[n,m]
Next, the UE 110 compresses the decomposition matrices U[n,m] and transmits at least one parameter associated with the decomposition matrices of the covariance matrices to the BS 121. In particular, uij[n,m] represents an entry in row i and column j of U[n,m]. The entries uij at different times (N number in total) and frequencies (M number in total) form matrices Uij.
The matrices Uij are compressed by: (1) being projected onto 2D time-frequency basis matrices; or (2) being represented as linear combination of 2D sinusoids matrices. More specifically, after compression, the compressed decomposition matrices can be represented as:
while Bkl is known parameter by both the UE 110 and the BS 121.
The UE 110 selects a subset Sij from the 2D time-frequency basis matrices (or from the 2D sinusoids matrices) and coefficients (cijkl:(k,l)∈Sij) corresponding to the subset, and then, as shown in
Then, the BS 121 reconstructs matrices Uij by Uij=Σ(k,l)∈S
Then, the BS 121 transmits subsequent PDSCHs to the UE 110 by using the precoders W.
In some implementations, the UE 110 calculates (or approximates) a decomposition of the covariance matrices Ψ[n,m] as:
Ψ[n,m]=UH[n,m]U[n,m]
Next, the UE 110 compresses the decomposition matrices U[n,m] and transmits at least one parameter associated with the decomposition matrices of the covariance matrices to the BS 121. In particular, uij[n,m] represents an entry in row i and column j of U[n,m]. The entries uij at different frequencies (M number in total) in time n form vector Uij[n].
The vectors Uij[n] are compressed by: (1) being projected onto 1D frequency basis vectors; or (2) being represented as linear combination of 1D sinusoids vectors.
More specifically, after compression, the compressed decomposition vector can be represented as:
U
ij[n]=Σk=1Mcijk[n]bk
while bk is known parameter by both the UE 110 and the BS 121.
The UE 110 selects a subset Sij[n] from the 1D frequency basis vectors (or from the 1D sinusoids vectors) and coefficients (cijk[n]:k∈Sij[n]) corresponding to the subset, and then, as shown in
Then, the BS 121 reconstructs vectors Uij[n] by Uij[n]=Σk∈S
Then, the BS 121 transmits subsequent PDSCHs to the UE 110 by using the precoders W.
In some implementations, the UE 110 calculates (or approximates) a decomposition of the covariance matrices Ψ[n,m] as:
Ψ[n,m]=UH[n,m]U[n,m],
where U[n,m] can be further decomposed into two terms Uv[n,m] and Uf. Uv[n,m] is time-frequency varying and Uf is time-frequency fixed. For example, U[n,m] is a matrix product of Uv[n,m] and Uf, which means U[n,m]=Uv[n,m]Uf or UfU[n,m]=Uv[n,m].
In some implementations, Uv[n,m] includes singular vectors and Uf includes average of squared rooted singular values. In some implementations, Uv[n,m] includes squared rooted singular values and Uf includes average of singular vectors
Next, the UE 110 compresses the decomposition matrices Uv[n,m] and transmits at least one parameter associated with the decomposition matrices of the covariance matrices to the BS 121. In particular, uv,ij[n,m] represents an entry in row i and column j of Uv[n,m]. The entries uv,ij at different times (N number in total) and frequencies (M number in total) form matrices Uv,ij.
The matrices Uv,ij are compressed by: (1) being projected onto 2D time-frequency basis matrices; or (2) being represented as linear combination of 2D sinusoids matrices. More specifically, after compression, the compressed decomposition matrices can be represented as:
while Bkl is known parameter by both the UE 110 and the BS 121.
The UE 110 selects a subset Sij from the 2D time-frequency basis matrices (or from the 2D sinusoids matrices) and coefficients (cijkl:(k,l)∈Sij) corresponding to the subset, and then, as shown in
Then, the BS 121 reconstructs matrices Uv,ij by Uv,ij=Σ(k,l)∈S
Then, the BS 121 transmits subsequent PDSCHs to the UE 110 by using the precoders W.
In some implementations, the UE 110 calculates (or approximates) a decomposition of the covariance matrices Ψ[n,m] as:
Ψ[n,m]=UH[n,m]U[n,m],
where U[n,m] can be further decomposed into two terms Uv[n,m] and Uf. Uv[n,m] is time-frequency varying and Uf is time-frequency fixed. For example, U[n,m] is a matrix product of Uv[n,m] and Uf, which means U[n,m]=Uv[n,m]Uf or UfU[n,m]=Uv[n,m]. In some implementations, Uv[n,m] includes singular vectors and Uf includes average of squared rooted singular values. In some implementations, Uv[n,m] includes squared rooted singular values and Uf includes average of singular vectors.
Next, the UE 110 compresses the decomposition matrices Uv[n,m] and transmits at least one parameter associated with the decomposition matrices of the covariance matrices to the BS 121. In particular, uv,ij[n,m] represents an entry in row i and column j of Uv[n,m]. The entries uv,ij at different frequencies (M number in total) in time n form vector Uv,ij[n].
The vectors Uv,ij[n] are compressed by: (1) being projected onto 1D frequency basis vectors; or (2) being represented as linear combination of 1D sinusoids vectors. More specifically, after compression, the compressed decomposition vectors can be represented as:
U
v,ij[n]=Σk=1Mcijk[n]bk
while bk is known parameter by both the UE 110 and the BS 121.
The UE 110 selects a subset Sij[n] from the 1D frequency basis vectors (or from 1D sinusoids vectors) and coefficients (cijk[n]:k∈Sij[n]) corresponding to the subset, and then, as shown in
Then the BS 121 reconstructs vectors Uv,ij[n] by Uv,ij[n]=Σk∈S
Then, the BS 121 transmits subsequent PDSCHs to the UE 110 by using the precoders W.
It should be noted that the parameters associated with 2D time-frequency basis matrices projection, 1D frequency basis vectors projection, linear combination of 2D sinusoids matrices and linear combination of 1D sinusoids vectors may be calculated by algorithms of 2D-discrete Fourier transform (DFT), 2D-estimating signal parameter via rotational invariance techniques (ESPRIT), 2D-multiple signal classification (MUSIC), 1D-DFT, 1D-ESPRIT and 1D-MUSIC.
In step 504, the UE transmits the at least one first covariance matrix or at least one parameter associated with the least one first covariance matrix to the BS. In step 505, the BS receives the at least one first covariance matrix or at least one parameter associated with the least one first covariance matrix from the UE.
In step 506, the BS reconstructs at least one second covariance matrix according to the at least one first covariance matrix or the at least one parameter associated with the least one first covariance matrix. In step 507, the BS derives a precoder according to the at least one second covariance matrix. In step 508, the BS transmits PDSCHs to the UE by using the precoder.
Although the present invention has been described in connection with certain specific embodiments for instructional purposes, the present invention is not limited thereto. Accordingly, various modifications, adaptations, and combinations of various features of the described embodiments can be practiced without departing from the scope of the invention as set forth in the claims.
This application claims the benefit under 35 U.S.C. § 119 from U.S. provisional application Ser. No. 63/236,709, entitled “MIMO CSI exploiting time domain correlation,” filed on Aug. 25, 2021, the subject matter of which is incorporated herein by reference.
Number | Date | Country | |
---|---|---|---|
63236709 | Aug 2021 | US |