Rapidly adaptable channel equalizer

Claim 2

For use in a digital data transfer apparatus that includes an adaptive equalizer means for receiving, correcting, and outputting symbols, a method of determining correction coefficients for the adaptive equalizer means that includes the steps of: for every symbol output by the adaptive equalizer means, maintaining a first matrix R that is a N×N channel correlation matrix and a second matrix P that is a N×1 crosscorrelation vector between a received signal and transmitted symbols; for every k th symbol output by the adaptive equalizer means, where k is equal to or greater than two, determining a set of equalizer coefficients (C) for the adaptive equalizer means based upon matrices R and P; and applying the set of determined correction coefficients to the adaptive equalizer means so as to correct a subsequently received data symbol.

Claim 3

A method as set forth in claim 1 wherein the step of maintaining includes the steps of operating on the matrix R in accordance with the expressions: R(n+1)=r×R(n)+X(n)×X(n).sup.T, and P(n+1)=r×P(n)+d(n)×X(n), where X is a vector formed by received symbol samples, d(n) is a symbol, the value of which is either known beforehand or is detected, r has a value in the range of 0<r≦1, n is a time index, and T indicates a transpose of a matrix.

Claim 4

A method as set forth in claim 2 wherein only the diagonal matrix D and either an upper triangular matrix or a lower triangular matrix of the matrix R is maintained.

Claim 5

A method as set forth in claim 1 wherein the step of determining include a step of: solving the expression RC=P by the steps of, STEP A: partitioning the matrix R into three matrices S, D, and S H so that R=S×D×S.sup.H, where S is a triangular matrix of R and D is the diagonal matrix of R, the step of positioning including a step of solving the formulas: D(1)=R(1,1) ##EQU5## where 1≦j≦i-1 and 2≦i≦N, and ##EQU6## where 2≦i≦N; STEP B: defining a matrix Y=D×S H ×C, and solving the group of equations: S×Y=P in accordance with the formulas: Y(1)=P(1), and ##EQU7## where 2≦i≦N; and STEP C: determining the set of correction coefficients C in accordance with: S.sup.H ×C=D.sup.-1 ×Y by the use of the formulas: C(N)=Y(N)/D(N), and ##EQU8## where 1≦i≦N-1.

Claim 6

A method as set forth in claim 4, and further including a step of defining a NxN matrix U for maintaining the matrices S and D as: U=S×D-D+D.sup.-1 +S.sup.H -I, where I is a diagonal unit matrix.

Claim 7

A radio telephone comprising: means for receiving and for transmitting an RF signal that is modulated with digital data representative of speech; coding and decoding means for applying and extracting the digital data as data symbols, said coding and decoding means including, for a receive path, adaptive equalizer means for receiving, correcting, and outputting data symbols; and detector means for detecting the data symbols output by said adaptive equalizer means, said detector means including means for determining correction coefficients for the adaptive equalizer means, said correction coefficient determining means including means, responsive to every data symbol output by said adaptive equalizer means, for maintaining a first matrix R that is a N×N channel correlation matrix and a second matrix P that is a N×1 crosscorrelation vector between a received signal and transmitted symbols, said correction coefficient determining means further being responsive to every k th symbol output by said adaptive equalizer means for determining a set of correction coefficients (C) for the adaptive equalizer means based upon the N×N channel correlation matrix R, said detector means having an output coupled to an input of said adaptive equalizer means for applying the set of determined correction coefficients to the adaptive equalizer means so as to optimally correct a subsequently received data symbol, wherein k is equal to or greater than two.

Claim 8

A radio telephone comprising: means for receiving and for transmitting an RF signal that is modulated with digital data representative of speech; coding and decoding means for applying and extracting the digital data as data symbols, said coding and decoding means including, for a receive path, adaptive equalizer means for receiving, correcting, and outputting data symbols; and detector means for detecting the data symbols output by said adaptive equalizer means, said detector means including means for determining correction coefficients for the adaptive equalizer means, said correction coefficient determining means including means, responsive to every data symbol output by said adaptive equalizer means, for maintaining a first matrix R that is a N×N channel correlation matrix and a second matrix P that is a N×1 crosscorrelation vector between a received signal and transmitted symbols, said correction coefficient determining means further being responsive to every k th symbol output by said adaptive equalizer means for determining a set of correction coefficients (C) for the adaptive equalizer means based upon the N×N channel correlation matrix R, said detector means having an output coupled to an input of said adaptive equalizer means for applying the set of determined correction coefficients to the adaptive equalizer means so as to optimally correct a subsequently received data symbol, wherein k is equal to or greater than two.

Claim 9

A radio telephone as set forth in claim 6 wherein said correction coefficient determination means operates on the matrix R in accordance with the expressions: R(n+1)=r×R(n)+X(n)×X(n).sup.T, and P(n+1)=r×P(n)+d(n)×X(n), where X is a vector formed by received data symbol samples, d(n) is a data symbol sent, the value of which is either known beforehand or is detected, r has a value in the range of 0<r≦1, n is a time index, and T indicates a transpose of a matrix.

Claim 10

A radio telephone as set forth in claim 7 and further including memory means for storing the diagonal matrix D and either an upper triangular matrix or a lower triangular matrix of the matrix R.

Claim 11

A radio telephone as set forth in claim 6 wherein said correction coefficient determination means solves an expression RC=P by a first operation of: partitioning the matrix R into three matrices S, D, and S H so that R=S×D×S.sup.H, where S is a triangular matrix of R and D is the diagonal matrix of R, the operation of partitioning including the operations of solving the formulas: D(1)=R(1,1) ##EQU9## where 1≦j≦i-1 and 2≦i≦N, and ##EQU10## where 2≦i≦N; by an operation of defining a matrix Y=D×S.sup.H ×C, and solving the group of equations: S×Y=P in accordance with the formulas: Y(1)=P(1), and ##EQU11## where 2≦i≦N; and by an operation of determining the set of correction coefficients C in accordance with: S.sup.H ×C=D.sup.-1 ×Y by the use of the formulas: C(N)=Y(N)/D(N), and ##EQU12## where 1≦i≦N-1.

Claim 12

A radio telephone as set forth in claim 9 and further including memory means for storing a N×N matrix U so as to maintain the matrices S and D as: U=S×D-D+D.sup.-1 +S.sup.H -I, where I is a diagonal unit matrix.