# US 10,455,079 B2 · Effective crosstalk estimation in presence of clipping errors > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 10,455,079 B2 - Document ID: US10455079B2 - Status: Granted - Kind: B2 - Application number: 15/697,731 - Filed: 2017-09-07 - Published: 2019-03-07 - Granted: 2019-10-22 - Earliest claimed: 2017-09-07 - Assignees: Nokia Solutions and Networks Oy - Applicants: Nokia Solutions and Networks OY - Inventors: Carl J. Nuzman - Web page: https://patentgrep.com/patent/US10455079B2 ## CPC - H04M 3/007: {Access interface units for simultaneous transmission of speech and data, e.g. digital subscriber line [DSL] access interface units (DSL access multiplexers H04Q11/0478, and H04L12/2856)} - H04B 3/32: Reducing cross-talk, e.g. by compensating - H04M 3/18: with means for reducing interference {or noise}; with means for reducing effects due to line faults {with means for protecting lines} - H04B 7/0421: {utilizing implicit feedback, e.g. steered pilot signals} - H04B 7/0634: {Antenna weights or vector/matrix coefficients} - H04M: TELEPHONIC COMMUNICATION (circuits for controlling other apparatus via a telephone cable and not involving telephone switching apparatus G08) - H04B: TRANSMISSION ## Abstract A method of reducing crosstalk into a victim line from disturber lines in a communication system may include obtaining a pilot matrix and a measurement vector associated with the pilot matrix, the pilot matrix representing pilot signals transmitted by at least the disturber lines, the measurement vector being influenced by the pilot signals via the crosstalk; estimating estimated crosstalk coefficients of significant ones of the disturber lines by solving a reduced linear system using only non-clipped measurements in the measurement vector, the reduced linear system being derived from the pilot matrix; determining compensation coefficients to reduce the crosstalk based on the estimated crosstalk coefficients; and reducing crosstalk by one or more of pre-compensating signals transmitted on the victim line and post-compensating signals received on the victim line based on the compensation coefficients. ## Claims ### Claim 1 A method of reducing crosstalk into a victim line from disturber lines in a communication system, the method comprising: obtaining a pilot matrix and a measurement vector associated with the pilot matrix, the pilot matrix representing pilot signals transmitted by at least the disturber lines, the measurement vector being influenced by the pilot signals via the crosstalk; estimating crosstalk coefficients of significant ones of the disturber lines by solving a reduced linear system using only non-clipped measurements in the measurement vector, the reduced linear system being derived from the pilot matrix; determining compensation coefficients to reduce the crosstalk based on the crosstalk coefficients; and reducing crosstalk by one or more of pre-compensating signals transmitted on the victim line and post-compensating signals received on the victim line based on the compensation coefficients. ### Claim 2 (depends on 1) The method of claim 1, further comprising: determining which elements of the measurement vector are clipped measurements and which of the elements are the non-clipped measurements; and determining which of the disturber lines are insignificant disturber lines and significant disturber lines, wherein the estimating includes setting the crosstalk coefficients associated with the insignificant disturber lines to zero. ### Claim 3 (depends on 2) The method of claim 2, wherein the estimating comprises: determining reduced pilot matrices; and estimating the crosstalk coefficients by solving a least squares problem based on the reduced pilot matrices. ### Claim 4 (depends on 3) The method of claim 3, wherein the solving the least squares problem comprises: computing pseudo-inverse matrices; and multiplying the non-clipped measurements by the pseudo-inverse matrices to estimate the crosstalk coefficients. ### Claim 5 (depends on 2) The method of claim 2, wherein the estimating comprises: applying the pilot matrix to initial sparse ones of the crosstalk coefficients to generate predicted measurement values; overwriting clipped measurement values with the predicted measurement values to generate corrected measurement values; multiplying the corrected measurement values by a transpose of the pilot matrix to generate dense ones of the crosstalk coefficients; and overwriting ones of the dense ones of the crosstalk coefficients associated with the insignificant disturber lines to zero to obtain improved sparse ones of the crosstalk coefficients. ### Claim 6 (depends on 5) The method of claim 5, further comprising: determining the initial sparse ones of the crosstalk coefficients for one of a plurality of tones based on extrapolating or interpolating the improved sparse ones of the crosstalk coefficients associated with other tones of the plurality of tones. ### Claim 7 (depends on 6) The method of claim 6, wherein the other tones of the plurality of tones are tones having a lower frequency than the one of the plurality of tones. ### Claim 8 (depends on 4) The method of claim 4, wherein the estimating comprises: iteratively improving the crosstalk coefficients until a stop condition is satisfied. ### Claim 9 (depends on 8) The method of claim 8, wherein the estimating includes determining that the stop condition is satisfied when the crosstalk coefficients for a current iteration is sufficiently close to the crosstalk coefficients for a previous iteration. ### Claim 10 (depends on 1) The method of claim 1, wherein the obtaining includes obtaining the pilot matrix such that columns of the pilot matrix represent columns of a Fourier matrix or Walsh-Hadamard Matrix. ### Claim 11 A device comprising: one or more processors configured to reduce crosstalk into a victim line from disturber lines in a communication system by, obtaining a pilot matrix and a measurement vector associated with the pilot matrix, the pilot matrix representing pilot signals transmitted by at least the disturber lines, the measurement vector being influenced by the pilot signals via the crosstalk, estimating crosstalk coefficients of significant ones of the disturber lines by solving a reduced linear system using only non-clipped measurements in the measurement vector, the reduced linear system being derived from the pilot matrix, determining compensation coefficients to reduce the crosstalk based on the crosstalk coefficients, and reducing crosstalk by one or more of pre-compensating signals transmitted on the victim line and post-compensating signals received on the victim line based on the compensation coefficients. ### Claim 12 (depends on 11) The device of claim 11, wherein the one or more processors are further configured to, determine which elements of the measurement vector are clipped measurements and which of the elements are the non-clipped measurements; and determine which of the disturber lines are insignificant disturber lines and significant disturber lines, wherein the one or more processors are configured to estimate the crosstalk coefficients of significant ones of the disturber lines by setting the crosstalk coefficients associated with the insignificant disturber lines to zero. ### Claim 13 (depends on 12) The device of claim 12, wherein the one or more processors are configured to estimate the crosstalk values of significant ones of the disturber lines by, determining reduced pilot matrices; and estimating the crosstalk coefficients by solving a least squares problem based on the reduced pilot matrices. ### Claim 14 (depends on 12) The device of claim 12, wherein the one or more processors are configured to estimate the crosstalk values of significant ones of the disturber lines by, applying the pilot matrix to initial sparse ones of the crosstalk coefficients to generate predicted measurement values; overwriting clipped measurement values with the predicted measurement values to generate corrected measurement values; multiplying the corrected measurement values by a transpose of the pilot matrix to generate dense ones of the crosstalk coefficients; and overwriting ones of the dense ones of the crosstalk coefficients associated with the insignificant disturber lines to zero to obtain improved sparse ones of the crosstalk coefficients. ### Claim 15 (depends on 14) The device of claim 14, wherein the one or more processors are further configured to, determine the initial sparse ones of the crosstalk coefficients for one of a plurality of tones based on extrapolating or interpolating the improved sparse ones of the crosstalk coefficients associated with other tones of the plurality of tones. ### Claim 16 (depends on 15) The device of claim 15, wherein the other tones of the plurality of tones are tones having a lower frequency than the one of the plurality of tones. ### Claim 17 (depends on 14) The device of claim 14, wherein the one or more processors are configured to estimate the crosstalk values of significant ones of the disturber lines by, iteratively improving the crosstalk coefficients until a stop condition is satisfied. ### Claim 18 (depends on 17) The device of claim 17, wherein the one or more processors are configured to determine that the stop condition is satisfied when the crosstalk coefficients for a current iteration is sufficiently close to the crosstalk coefficients for a previous iteration. ### Claim 19 (depends on 11) The device of claim 11, wherein the one or more processors are configured to obtain the pilot matrix such that columns of the pilot matrix represent columns of a Fourier matrix or Walsh-Hadamard Matrix. ### Claim 20 A method of estimating crosstalk into a victim line from disturber lines in a communication system, the method comprising: obtaining a pilot matrix and a measurement vector associated with the pilot matrix, the pilot matrix representing pilot signals transmitted by at least the disturber lines, the measurement vector being influenced by the pilot signals via the crosstalk; and estimating crosstalk coefficients of significant ones of the disturber lines by solving a reduced linear system using only non-clipped measurements in the measurement vector, the reduced linear system being derived from the pilot matrix.