# US 7,613,248 B2 · Signal processing with channel eigenmode decomposition and channel inversion for MIMO systems > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 7,613,248 B2 - Document ID: US7613248B2 - Status: Granted - Kind: B2 - Application number: 10/179,442 - Filed: 2002-06-24 - Published: 2003-12-25 - Granted: 2009-11-03 - Earliest claimed: 2002-06-24 - Assignees: QUALCOMM Incorporated - Inventors: John W. Ketchum; Jay R. Walton - Web page: https://patentgrep.com/patent/US7613248B2 ## CPC - H04B 7/0417: Feedback systems - H04B 7/0626: {Channel coefficients, e.g. channel state information [CSI]} - H04L 1/0618: {Space-time coding} - H04L 25/0202: {Channel estimation} - H04L 25/03343: {Arrangements at the transmitter end} - H04L 25/03834: {using pulse shaping} - H04B 7/0439: {utilizing channel inversion} - H04B 7/0443: {utilizing "waterfilling" technique} - H04B: TRANSMISSION - H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M) ## Abstract Techniques for processing a data transmission at a transmitter and receiver, which use channel eigen-decomposition, channel inversion, and (optionally) “water-pouring”. At the transmitter, (1) channel eigen-decomposition is performed to determine eigenmodes of a MIMO channel and to derive a first set of steering vectors, (2) channel inversion is performed to derive weights (e.g., one set for each eigenmode) used to minimize ISI distortion, and (3) water-pouring may be performed to derive scaling values indicative of the transmit powers allocated to the eigenmodes. The first set of steering vectors, weights, and scaling values are used to derive a pulse-shaping matrix, which is used to precondition modulation symbols prior to transmission. At the receiver, channel eigen-decomposition is performed to derive a second set of steering vectors, which are used to derive a pulse-shaping matrix used to condition received symbols such that orthogonal symbol streams are recovered. ## Claims ### Claim 1 In a multiple-input multiple-output (MIMO) communication system, a method for processing data for transmission over a MIMO channel, comprising: processing data in accordance with a particular processing scheme to provide a plurality of streams of modulation symbols; deriving a pulse-shaping matrix based on an estimated channel response matrix of the MIMO channel and a plurality of scaling values used to adjust transmit rower for the MIMO channel, which are based on a plurality of matrices of singular values, wherein the pulse-shaping matrix comprises a plurality of time-ordered sequences of values; and preconditioning, according to at least a convolution operation, the plurality of modulation symbol streams, based on the pulse-shaping matrix to provide a plurality of streams of preconditioned symbols for transmission over the MIMO channel. ### Claim 2 (depends on 1) The method of claim 1, further comprising: deriving a plurality of weights based on an estimated channel response matrix for the MIMO channel, wherein the weights are used to invert a frequency response of the MIMO channel, and wherein the pulse-shaping matrix is further derived based on the weights. ### Claim 3 (depends on 2) The method of claim 2, further comprising: decomposing the estimated channel response matrix to obtain a plurality of matrices of eigen-vectors and the plurality of matrices of singular values, and wherein the weights are derived based on the matrices of singular values and the pulse-shaping matrix is further derived based on the matrices of eigen-vectors. ### Claim 4 (depends on 2) The method of claim 2, wherein the estimated channel response matrix is descriptive of a plurality of eigenmodes of the MIMO channel. ### Claim 5 (depends on 4) The method of claim 4, wherein one set of weights is derived for each eigenmode used for data transmission and wherein the weights in each set are derived to invert the frequency response of the corresponding eigenmode. ### Claim 6 (depends on 4) The method of claim 4, wherein the scaling values are used to adjust transmit powers for the eigenmodes of the MIMO channel. ### Claim 7 (depends on 6) The method of claim 6, wherein the scaling values are derived based on water-pouring analysis. ### Claim 8 (depends on 3) The method of claim 3, wherein the estimated channel response matrix is provided in the frequency domain and is decomposed in the frequency domain. ### Claim 9 (depends on 3) The method of claim 3, wherein the estimated channel response matrix is decomposed using channel eigen-decomposition. ### Claim 10 (depends on 4) The method of claim 4, wherein eigenmodes associated with transmission capabilities below a particular threshold are not used for data transmission. ### Claim 11 (depends on 3) The method of claim 3, wherein the singular values in each matrix are sorted based on their magnitude. ### Claim 12 (depends on 4) The method of claim 4, wherein the singular values in each matrix are randomly ordered such that the eigenmodes of the MIMO channel are associated with approximately equal transmission capabilities. ### Claim 13 (depends on 1) The method of claim 1, wherein the pulse-shaping matrix comprises a plurality of time sequences of values representing a response of the MIMO channel at discrete frequencies, and wherein the preconditioning is performed by multiplying a plurality of streams of transformed modulation symbols with the pulse-shaping matrix. ### Claim 14 (depends on 1) The method of claim 1, the MIMO channel having eigenmodes having greater transmission capability and eigenmodes having lesser transmission capabilities, wherein the pulse-shaping matrix is derived to maximize capacity by allocating more transmit power to the eigenmodes of the MIMO channel having greater transmission capabilities. ### Claim 15 (depends on 1) The method of claim 1, wherein the pulse-shaping matrix is derived to provide approximately equal received signal-to-noise-and-interference ratios (SNRs) for the plurality of modulation symbol streams at the receiver. ### Claim 16 (depends on 1) The method of claim 1, wherein the particular processing scheme defines a separate coding and modulation scheme for each modulation symbol stream. ### Claim 17 (depends on 1) The method of claim 1, wherein the particular processing scheme defines a common coding and modulation scheme for all modulation symbol streams. ### Claim 18 In a multiple-input multiple-output (MIMO) communication system, a method for processing data for transmission over a MIMO channel, comprising: processing data in accordance with a particular processing scheme to provide a plurality of streams of modulation symbols; obtaining an estimated channel response matrix for the MIMO channel; decomposing the estimated channel response matrix to obtain a plurality of matrices of eigen-vectors and a plurality of matrices of singular values; deriving a plurality of weights based on the matrices of singular values, wherein the weights are used to invert the frequency response of the MIMO channel; deriving a plurality of scaling values based on the matrices of singular values, wherein the scaling values are used to adjust transmit powers for eigenmodes of the MIMO channel; deriving a pulse-shaping matrix based on the matrices of eigen-vectors the plurality of scaling values, and the weights, wherein the pulse-shaping matrix comprises a plurality of sequences of time-domain values; and preconditioning, according to at least a convolution operation, the plurality of streams of modulation symbols in the time domain, based on the pulse-shaping matrix to provide a plurality of streams of preconditioned symbols for transmission over the MIMO channel. ### Claim 19 A computer readable media memory storing instructions to be executed by a digital signal processing device (DSPD) capable of interpreting the instructions to perform the steps of: processing data in accordance with a particular processing scheme to provide a plurality of streams of modulation symbols; deriving a pulse-shaping matrix based on an estimated channel response matrix of the MIMO channel and a plurality of transmit power scaling values, which are based on a plurality of matrices of singular values, wherein the pulse-shaping matrix comprises a plurality of sequences of time-domain values; and precondition, according to at least a convolution operation, the plurality of streams of modulation symbols in the time domain, based on the pulse-shaping matrix to provide a plurality of streams of preconditioned symbols for transmission over the MIMO channel. ### Claim 20 In a multiple-input multiple-output (MIMO) communication system, a method for processing a data transmission received via a MIMO channel, comprising: obtaining an estimated channel response matrix for the MIMO channel; decomposing the estimated channel response matrix to obtain a plurality of matrices of eigen-vectors; deriving a time domain pulse-shaping matrix based on the matrices of eigen-vectors; and conditioning, according to at least a convolution operation in the time domain, a plurality of streams of received symbols, based on the time domain pulse-shaping matrix to provide a plurality of streams of recovered symbols which are estimates of preconditioned modulation symbols transmitted for the data transmission, wherein the preconditioned modulation symbols are preconditioned at a transmitter prior to transmission over the MIMO channel, and are based on a plurality of scaling values used to adjust transmit power for the MIMO channel, and wherein the plurality of scaling values are based on a plurality of matrices of singular values. ### Claim 21 (depends on 20) The method of claim 20, wherein the conditioning orthogonalizes a plurality of streams of modulation symbols transmitted over the MIMO channel. ### Claim 22 (depends on 20) The method of claim 20, further comprising: demodulating the plurality of recovered symbol streams in accordance with one or more demodulation schemes to provide a plurality of demodulated data streams; and decoding the plurality of demodulated data streams in accordance with one or more decoding schemes to provide decoded data. ### Claim 23 (depends on 20) The method of claim 20, further comprising: deriving channel state information (CSI) comprised of the estimated channel response matrix for the MIMO channel; and sending the CSI back to the transmitter. ### Claim 24 In a multiple-input multiple-output (MIMO) communication system, a method for processing a data transmission received via a MIMO channel, comprising: obtaining an estimated channel response matrix for the MIMO channel; decomposing the estimated channel response matrix to obtain a plurality of matrices of eigen-vectors; deriving a time domain pulse-shaping matrix based on the matrices of eigen-vectors; conditioning, according to at least a convolution operation in the time domain, a plurality of streams of received symbols, based on the time domain pulse-shaping matrix to provide a plurality of streams of recovered symbols which are estimates of preconditioned modulation symbols transmitted for the data transmission, wherein the preconditioned modulation symbols are preconditioned at a transmitter prior to transmission over the MIMO channel, and are based on a plurality of scaling values used to adjust transmit power for the MIMO channel, and wherein the plurality of scaling values are based on a plurality of matrices of singular values; demodulating the plurality of recovered symbol streams in accordance with one or more demodulation schemes to provide a plurality of demodulated data streams; and decoding the plurality of demodulated data streams in accordance with one or more decoding schemes to provide decoded data. ### Claim 25 A computer readable media memory storing instructions to be executed by a digital signal processing device (DSPD) capable of interpreting the instructions to perform the steps of: obtaining an estimated channel response matrix for a multiple-input multiple output (MIMO) channel used for a data transmission; decomposing the estimated channel response matrix to obtain a plurality of matrices of eigen-vectors; deriving a time domain pulse-shaping matrix based on the matrices of eigen-vectors; and conditioning, according to at least a convolution operation in the time domain, a plurality of streams of received symbols, based on the time domain pulse-shaping matrix to provide a plurality of streams of recovered symbols which are estimates of preconditioned modulation symbols transmitted for the data transmission, wherein the preconditioned modulation symbols are preconditioned at a transmitter prior to transmission over the MIMO channel, and are based on a plurality of scaling values used to adjust transmit power for the MIMO channel, and wherein the plurality of scaling values are based on a plurality of matrices of singular values. ### Claim 26 A transmitter unit in a multiple-input multiple-output (MIMO) communication system, comprising: a transmit (TX) data processor operative to process data in accordance with a particular processing scheme to provide a plurality of streams of modulation symbols; and a TX MIMO processor operative to derive a pulse-shaping matrix based on an estimated response of a MIMO channel, and to precondition, according to at least a convolution operation, the plurality of modulation symbol streams, based on the pulse-shaping matrix and on a plurality of scaling values used to adjust transmit power for the MIMO channel, which are based on a plurality of matrices of singular values, and to provide a plurality of streams of preconditioned symbols for transmission over the MIMO channel, wherein the pulse-shaping matrix comprises a plurality of sequences of time-domain values, and wherein the preconditioning is performed in the time domain by convolving the streams of modulation symbols with the pulse-shaping matrix. ### Claim 27 (depends on 26) The transmitter unit of claim 26, wherein the TX MIMO processor is further operative to derive a plurality of weights based on an estimated channel response matrix for the MIMO channel, wherein the weights are used to invert the frequency response of the MIMO channel, and wherein the pulse-shaping matrix is derived based in part on the weights. ### Claim 28 (depends on 27) The transmitter unit of claim 27, wherein the TX MIMO processor is further operative to decompose the estimated channel response matrix to obtain a plurality of matrices of eigen-vectors and the plurality of matrices of singular values, and wherein the weights are derived based on the matrices of singular values and the pulse-shaping matrix is further derived based on the matrices of eigen-vectors. ### Claim 29 (depends on 27) The transmitter unit of claim 27, the MIMO channel having plural eigenmodes, wherein the plurality of scaling values are used to adjust transmit powers for the plural eigenmodes of the MIMO channel. ### Claim 30 (depends on 29) The transmitter unit of claim 29, wherein the scaling values are derived based on water-pouring analysis. ### Claim 31 An apparatus in a multiple-input multiple-output (MIMO) communication system, comprising: means for processing data in accordance with a particular processing scheme to provide a plurality of streams of modulation symbols; means for deriving a pulse-shaping matrix based on an estimated channel response matrix of a MIMO channel and a plurality of scaling values used to adjust transmit rower for the MIMO channel, which are based on a plurality of matrices of singular values, wherein the pulse-shaping matrix comprises a plurality of sequences of time-domain values; and means for preconditioning, according to at least a convolution operation, the plurality of modulation symbol streams in the time domain, based on the pulse-shaping matrix to provide a plurality of streams of preconditioned symbols for transmission over the MIMO channel. ### Claim 32 A digital signal processor comprising: means for processing data in accordance with a particular processing scheme to provide a plurality of streams of modulation symbols; means for deriving a pulse-shaping matrix based on an estimated channel response matrix of a MIMO channel and a plurality of scaling values used to adjust transmit rower for the MIMO channel, which are based on a plurality of matrices of singular values, wherein the pulse-shaping matrix comprises a plurality of sequences of time-domain values; and means for preconditioning, according to at least a convolution operation, the plurality of modulation symbol streams in the time domain, based on the pulse-shaping matrix to provide a plurality of streams of preconditioned symbols for transmission over the MIMO channel. ### Claim 33 A receiver unit in a multiple-input multiple-output (MIMO) communication system, comprising: an RX MIMO processor operative to obtain an estimated channel response matrix for a MIMO channel used for a data transmission, decompose the estimated channel response matrix to obtain a plurality of matrices of eigen-vectors, derive a time domain pulse-shaping matrix based on the matrices of eigen-vectors, and condition, according to at least a convolution operation in the time domain, a plurality of streams of received symbols, based on the pulse-shaping matrix to obtain a plurality of streams of recovered symbols which are estimates of preconditioned modulation symbols transmitted over the MIMO channel, wherein the modulation symbols were preconditioned at a transmitter prior to transmission over the MIMO channel and are based on a plurality of scaling values used to adjust transmit power for the MIMO channel, and wherein the plurality of scaling values are based on a plurality of matrices of singular values, and an RX data processor operative to process the plurality of recovered symbol streams in accordance with a particular processing scheme to provide decoded data. ### Claim 34 An apparatus in a multiple-input multiple-output (MIMO) communication system, comprising: means for obtaining an estimated channel response matrix for a MIMO channel used for a data transmission; means for decomposing the estimated channel response matrix to obtain a plurality of matrices of eigen-vectors; means for deriving a time domain pulse-shaping matrix based on the matrices of eigen-vectors; and means for conditioning, according to at least a convolution operation in the time domain, a plurality of streams of received symbols, based on the time domain pulse-shaping matrix, to provide a plurality of streams of recovered symbols which are estimates of preconditioned modulation symbols transmitted for the data transmission, wherein the preconditioned modulation symbols are preconditioned at a transmitter prior to transmission over the MIMO channel, and are based on a plurality of scaling values used to adjust transmit power for the MIMO channel, and wherein the plurality of scaling values are based on a plurality of matrices of singular values. ### Claim 35 A digital signal processor comprising: means for obtaining an estimated channel response matrix for a multiple-input multiple-output (MIMO) channel used for a data transmission; means for decomposing the estimated channel response matrix to obtain a plurality of matrices of eigen-vectors; means for deriving a time domain pulse-shaping matrix based on the matrices of eigen-vectors; and means for conditioning, according to at least a convolution operation in the time domain, a plurality of streams of received symbols, based on the time domain pulse-shaping matrix, to provide a plurality of streams of recovered symbols which are estimates of preconditioned modulation symbols transmitted for the data transmission, wherein the preconditioned modulation symbols are preconditioned at a transmitter prior to transmission over the MIMO channel, and are based on a plurality of scaling values used to adjust transmit power for the MIMO channel, and wherein the plurality of scaling values are based on a plurality of matrices of singular values. ### Claim 36 (depends on 1) The method of claim 1, wherein the pulse-shaping matrix comprises a plurality of sequences of values representing a response of the MIMO channel at discrete points in time, and wherein the preconditioning is performed in the time domain by scaling the plurality of streams at the discrete points in time.