Method and apparatus for estimating parameters of a communication system using antenna arrays and spatial processing

Claim 2

In a communication system comprising at least one remote terminal and a base station, the base station having an antenna array of m antenna elements, the communication system having a set of one or more parameters, said set consisting of a first set of parameters already estimated and a second set of parameters to be estimated, a method for determining an estimate of one or more parameters of the second set, the method comprising: a) transmitting a first signal of known characteristics; b) receiving said first signal at the base station as a first plurality of m received signals at m corresponding antenna elements; c) determining a cost function that is related to the at least one or more parameters of the second set, said determining including: i) forming a copy signal from samples of the signals of the first plurality of received signals using a weight vector, ii) if the first set is not empty, obtaining estimates of one or more parameters of the first set, and iii) using the known characteristics of the first signal; and d) selecting as the estimate of the one or more parameters of the second set, the values of the one or more parameters of the second set that minimize the cost function.

Claim 3

The method of claim 1 wherein the first signal is a subset of a SYNCH burst and the known characteristics include the format of the subset of the SYNCH burst.

Claim 4

The method of claim 2 wherein the subset is the SYNCH burst.

Claim 5

The method of claim 2 wherein the subset is one or more prescribed fixed portions of the SYNCH signal.

Claim 6

The method of claim 1 wherein said first set is empty and said second set includes the time alignment of said system.

Claim 7

The method of claim 5 wherein said cost function is a quadratic cost function of the copy signal and a reference signal determined from the known characteristics of the first signal.

Claim 8

The method of claim 6 wherein said copy signal is of the amplitudes of samples of said received signals, and the reference signal is related to the amplitude of samples of said subset.

Claim 9

The method of claim 6 wherein said copy signal is of squared amplitudes of samples of said received signals, and the reference signal is related to the squared amplitude of samples of said subset.

Claim 10

In a communication system comprising at least one remote terminal and a base station, the base station having an antenna array of m antenna elements, the communication system having a set of one or more parameters, a method for determining an estimate of one or more parameters of the set, the method comprising: a) transmitting a first signal of known characteristics; b) receiving said first signal at the base station as a first plurality of m received signals at m corresponding antenna elements; c) determining a cost function that is related to the at least one or more parameters of the set, said cost function being a function of (i) a copy signal related to samples of the signals of the first plurality of received signals, and (ii) a reference signal determined from the known characteristics of the first signal; and d) selecting as the estimate of the one or more parameters of the set, the values of the one or more parameters of the set that minimize the cost function.

Claim 11

In a communication system comprising at least one remote terminal and a base station, the base station having an antenna array of m antenna elements, the communication system having a set of one or more parameters, a method for determining an estimate of one or more parameters of the set, the method comprising: a) transmitting a first signal of known characteristics; b) receiving said first signal at the base station as a first plurality of m received signals at m corresponding antenna elements; c) determining a cost function that is related to the at least one or more parameters of the set, said cost function being a function of (i) a copy signal related to samples of the signals of the first plurality of received signals, and (ii) a reference signal determined from the known characteristics of the first signal; and d) selecting as the estimate of the one or more parameters of the set, the values of the one or more parameters of the set that minimize the cost function.

Claim 12

The method of claim 9 wherein the first signal is a subset of a SYNCH burst and the known characteristics include the format of the subset of the SYNCH burst.

Claim 13

The method of claim 9 wherein said cost function is a quadratic cost function.

Claim 14

The method of claim 10 wherein said one or more parameters is the time alignment, the copy signal is of the amplitudes of said received signals, and the reference signal is related to the amplitude of said subset.

Claim 15

The method of claim 10 wherein said one or more parameters is the time alignment, said copy signal is of the squared amplitudes of said received signals, and the reference signal is related to the amplitude of said subset.

Claim 16

In a communications system comprising at least one remote terminal and at least one base station, each of said at least one base stations having m antenna elements, a method for estimating the time alignment of the communications system, the method comprising: a) transmitting a first signal of known characteristics; b) receiving said first signal as a first plurality of m received signals at m corresponding antenna elements; c) calculating a minimum cost function for each time alignment of a set of time alignments, said minimum cost function calculating comprising determining the weight vector that minimizes a cost function of the error between a reference signal and a copy signal, the reference signal determined from the known characteristics of said first signal and the copy signal being of samples of the signals of said first plurality of received signals, the copy signal calculated using said weight vector, the copy signal and the reference signal shifted in time relative to each other by said each time alignment, the minimum cost function being the cost function using said calculated weight vector; and d) selecting the time alignment estimate as the time alignment of said set of time alignments which has the lowest minimum cost function.

Claim 17

In a communications system comprising at least one remote terminal and at least one base station, each of said at least one base stations having m antenna elements, a method for estimating the time alignment of the communications system, the method comprising: a) transmitting a first signal of known characteristics; b) receiving said first signal as a first plurality of m received signals at m corresponding antenna elements; c) calculating a minimum cost function for each time alignment of a set of time alignments, said minimum cost function calculating comprising determining the weight vector that minimizes a cost function of the error between a reference signal and a copy signal, the reference signal determined from the known characteristics of said first signal and the copy signal being of samples of the signals of said first plurality of received signals, the copy signal calculated using said weight vector, the copy signal and the reference signal shifted in time relative to each other by said each time alignment, the minimum cost function being the cost function using said calculated weight vector; and d) selecting the time alignment estimate as the time alignment of said set of time alignments which has the lowest minimum cost function.

Claim 18

The method of claim 14 wherein the first signal is a subset of a SYNCH burst and the known characteristics include the format of the subset of the SYNCH burst.

Claim 19

The method of claim 14 wherein said cost function is a quadratic cost function.

Claim 20

The method of claim 16 wherein the copy signal is of the amplitudes of samples of said received signals, and the reference signal is related to the amplitude of said subset.

Claim 21

The method of claim 16 wherein the copy signal is of samples of the amplitudes of said received signals, and the reference signal is related to the amplitude squared of said subset.

Claim 22

The method of claim 18 wherein the minimization of the cost function is by use of a pseudoinverse process.

Claim 23

The method of claim 18 wherein the minimization of the cost function is by use of a conjugate gradient method.

Claim 24

The method of claim 18 wherein adaptive normalization is used at each step of the minimization.

Claim 25

The method of claim 1 wherein said first set includes time alignment and said second set includes the frequency offset of said system.

Claim 26

The method of claim 22 wherein said cost function is a function of the error between said copy signal and a reference signal, the reference signal formed from samples of said first signal.

Claim 27

The method of claim 23 wherein said cost function includes a correction of the timing of the copy signal relative to the timing of the samples of said first signal, said correction using the already determined estimate of time alignment in said first set.

Claim 28

The method of claim 23 wherein said cost function is a quadratic cost function of said error.

Claim 29

The method of claim 24 wherein said cost function is a quadratic cost function of said error.

Claim 30

The method of claim 25 wherein the minimization of the cost function is by use of a pseudoinverse process.

Claim 31

The method of claim 25 wherein the minimization of the cost function is by use of a conjugate gradient method.

Claim 32

The method of claim 26 wherein the minimization of the cost function is by use of a pseudoinverse process.

Claim 33

The method of claim 26 wherein the minimization of the cost function is by use of a conjugate gradient method.

Claim 34

The method of claim 25 wherein adaptive normalization is used at each step of the minimization.

Claim 35

The method of claim 25 wherein the reference signal is a frequency-offset replica of samples of said first signal, the shift being one of a set of frequency offsets.

Claim 36

The method of claim 32 wherein initially the offsets of said set of frequency offsets are separated by a frequency interval, and wherein said step of selecting comprises: a) determining the weight vector and frequency offset that minimizes said cost function, said weight vector the minimizing weight vector, said frequency offset the minimizing frequency offset; b) if the desired accuracy has not been obtained, repeating step d) at least once, with a new set of frequency offsets, the offsets of the new set centered around the most recent previous minimizing frequency offset, the offsets of the new set separated by a new interval less than the most recent previous frequency interval, said repeating until the desired accuracy has been obtained; and c) choosing as the estimate of the estimate of the frequency offset the last obtained minimizing frequency offset.

Claim 37

The method of claim 33 wherein said second set includes the weight vector and wherein said step of selecting includes choosing as the estimate of the weight vector the last obtained minimizing weight vector.

Claim 38

The method of claim 9 wherein said set includes the frequency offset of said system.

Claim 39

The method of claim 35 wherein said cost function is a function of the error between said copy signal and the reference signal, the reference signal formed from samples of said first signal.

Claim 40

The method of claim 36 wherein said cost function is a quadratic cost function of said error.

Claim 41

The method of claim 37 wherein the minimization of the cost function is by use of a pseudoinverse process.

Claim 42

The method of claim 37 wherein the minimization of the cost function is by use of a conjugate gradient method.

Claim 43

The method of claim 37 wherein adaptive normalization is used at each step of the minimization.

Claim 44

The method of claim 37 wherein the reference signal is a frequency-offset replica of samples of said first signal, the shift being one of a set of frequency offsets.

Claim 45

The method of claim 40 wherein initially the offsets of said set of frequency offsets are separated by a frequency interval, and wherein said step of selecting comprises: a) determining the weight vector and frequency offset that minimizes said cost function, said weight vector the minimizing weight vector, said frequency offset the minimizing frequency offset; b) if the desired accuracy has not been obtained, repeating step d) at least once, with a new set of frequency offsets, the offsets of the new set centered around the most recent previous minimizing frequency offset, the offsets of the new set separated by a new interval less than the most recent previous frequency interval, said repeating until the desired accuracy has been obtained; and c) choosing as the estimate of the frequency offset the last obtained minimizing frequency offset.

Claim 46

The method of claim 42 wherein said second set includes the weight vector and wherein said step of selecting includes choosing as the estimate of the weight vector the last obtained minimizing weight vector.

Claim 47

The method of claim 1 wherein the first set is empty and the second set includes the time alignment of the system.

Claim 48

The method of claim 1 wherein the first set is empty, the second set includes the time alignment of the system, said first signal modulated by a modulation scheme with a finite alphabet property, and the known characteristics including the modulation scheme.

Claim 49

The method of claim 45 wherein the symbols of the finite alphabet all differ in phase.

Claim 50

The method of claim 46 wherein the modulation scheme is DPSK, DPSK representable by a set of differential constellation points in the complex plane.

Claim 51

The method of claim 47 wherein said received signals and said copy signal are representable by complex values, the cost function is a monotonic function of the errors between a time shifted differential signal and the closest ideal differential signal to said time shifted differential signal, the differential signal defined as having a phase difference between subsequent samples equal to the phase difference between subsequent samples of the copy signal, said samples separated by the baud period, said error at any point in time being the distance in the complex plane between the differential signal and the nearest differential constellation point of said modulation scheme, said time shift by a predefined amount, said minimization determining the time shift that minimizes said cost function, the estimate of time alignment being the time shift that minimizes said cost function.

Claim 52

The method of claim 48 wherein said cost function is the sum of the squares of the errors.

Claim 53

In a communications system comprising at least one remote terminal and a base station, the base stations having a plurality of antenna elements, in which successful communication between the base station and a remote terminal requires that common channel timing and frequency be established, a method for establishing common channel timing and frequency by estimating a second set of parameters from a first set of estimated parameters, the method comprising: a) transmitting a modulated signal with known finite alphabet properties, said modulated signal modulated by a DPSK modulation scheme; b) receiving the transmitted signal at the plurality of antenna elements and producing a down-converted complex representation of each received signal as a received signal vector; c) estimating the transmitted signal using a signal copy operation with an initial spatial weight vector; d) creating a time sampled representation, b 0 (n), of s 0 (t); e) forming a differential stream, d.sub.0 (n)=b.sub.0 (n)/b.sub.0 (n-L) ∠d.sub.0,(n)=∠b.sub.0 (n)-∠b.sub.0 (n-L) where L is the baud interval and ∠(.) represents a phase angle of an associated argument; f) computing a time alignment squared error e A 2 (n)=|d 0 (n)-d 0ideal (n)| 2 as e A 2 (n)=(|x Re (n)|-1/√2) 2 +(|jx Im (n)|-1/√2) 2 where d 0idea (n) is the closest ideal complex value (±∠π/4 or ±∠3π/4) to d 0 (n), and b 0 (n)b 0 *(n-L)!=x Re (n)+jx Im (n); and g) minimizing a cost function J x with respect to offset index x where ##EQU2## and N is the total number of samples in d 0 (n), selecting the estimated time offset parameter to be the value of x that minimizes J x .

Claim 54

In a communications system comprising at least one remote terminal and a base station, the base stations having a plurality of antenna elements, in which successful communication between the base station and a remote terminal requires that common channel timing and frequency be established, a method for establishing common channel timing and frequency by estimating a second set of parameters from a first set of estimated parameters, the method comprising: a) transmitting a modulated signal with known finite alphabet properties, said modulated signal modulated by a DPSK modulation scheme; b) receiving the transmitted signal at the plurality of antenna elements and producing a down-converted complex representation of each received signal as a received signal vector; c) estimating the transmitted signal using a signal copy operation with an initial spatial weight vector; d) creating a time sampled representation, b 0 (n), of s 0 (t); e) forming a differential stream, d.sub.0 (n)=b.sub.0 (n)/b.sub.0 (n-L) ∠d.sub.0,(n)=∠b.sub.0 (n)-∠b.sub.0 (n-L) where L is the baud interval and ∠(.) represents a phase angle of an associated argument; f) computing a time alignment squared error e A 2 (n)=|d 0 (n)-d 0ideal (n)| 2 as e A 2 (n)=(|x Re (n)|-1/√2) 2 +(|jx Im (n)|-1/√2) 2 where d 0idea (n) is the closest ideal complex value (±∠π/4 or ±∠3π/4) to d 0 (n), and b 0 (n)b 0 *(n-L)!=x Re (n)+jx Im (n); and g) minimizing a cost function J x with respect to offset index x where ##EQU2## and N is the total number of samples in d 0 (n), selecting the estimated time offset parameter to be the value of x that minimizes J x .

Claim 55

The method of claim 50 further comprising steps for a frequency offset parameter estimation, the steps comprising: a) applying the estimated offset parameter as a correction to d 0 (n) and d 0ideal (n) and producing corresponding time aligned signals d' 0 (n) and d' 0ideal (n), respectively, and corresponding phase error of e P (n)=∠d' 0 (n)-∠d' 0ideal (n); b) minimizing a phase error cost function, J P , with respect to ∠d' 0 (n) by rotating ∠d' 0 (n), where J P is an average of e P (n) over all samples to obtain ∠d' 0min (n), a phase offset that minimizes J P ; and c) computing a frequency offset by dividing phase angle ∠d' 0min (n) by a time increment corresponding to increments of time index n.