US 2008/0175339 A1 · Method for the Flexible Demodulation of Estimated Complex Symbols

Abstract

A method for soft demapping of complex-valued symbols y(k) (k=1, 2, . . . , K) by calculation of a Logarithm of Likelihood Ratio (LogLR) such that: Log   LR  ( b I , q ) = ln  ∑ λ  ∈ S l , q ( 1 )  exp ( - 1 2  β   y ~  ( k ) - λ  2 ) ∑ λ  ∈ S q , l ( 0 )  exp ( - 1 2  β   y ~  ( k ) - λ  2 ) wherein bI,q is the qth coded bit of an in-phase part of a transmitted signal, β is an asymptotic Signal-to-Interference-plus-Noise Ratio, {tilde over (y)}(k) is a standardized symbol associated with y(k), λ is a reference symbol of QAM constellation associated with a M-QAM modulation, S(1)I,q and S(0)I,q are complex plane partitions that bring together, respectively, all reference symbols λ having a “1” at the position (I, q) and all reference symbols λ having a “0” at the position (I, q), a formula similar to LogLR being given for the bit bQ,p, pth coded bit of a quadrature part of the transmitted signal. The method applies to code division multiple access transmission systems.

Claims19

Dependent claims

1-8. (canceled)

A method for soft demapping, by Logarithm of Likelihood Ratio calculation, of complex-valued symbols (y(1), y(2), . . . , y(k), . . ., y(K)) corresponding to transmitted complex symbols (a(1), a(2), . . . , a(k), . . . , a(K)) contained in a signal transmitted by a code division multiple access multi-carrier transmitter and constituted by binary data coded by convolutional coding and quadrature modulated according to a M-QAM modulation,

wherein M is an integer equal to 2m and m is an integer greater than or equal to 1, a complex symbol transmitted a(k) of rank k, k=1, 2, . . . , K being spread by a spreading code associated with a user of rank k, the binary data coded by convolutional coding constituting, for the user of rank k, a sequence of bits {bI,1;
bI,2;
. . . , bI,q;
. . . , bI,r;
bQ,1;
bQ,2;
. . . , bQ,p;
. . . , bQ,s}, wherein the bit bI,q is the qth coded bit of an in-phase part of the transmitted signal and the bit bQ,p is the pth coded bit of a quadrature part of the transmitted signal, wherein the method includes a Logarithm of the Likelihood Ratio LogLR (bI,q) calculation of the bit bI,q and a Logarithm of the Likelihood Ratio LogLR (bQ,p) calculation of the bit bQ,p so that:
Log


LR

(
bI
,
q
)
=
ln

∑
λ

∈
S
I
,
q
(
1
)


exp
(
-
1
2

β


y
~

(
k
)
-
λ

2
)
∑
λ

∈
S
I
,
q
(
0
)


exp
(
-
1
2

β


y
~

(
k
)
-
λ

2
)
and
Log


LR

(
bQ
,
p
)
=
ln

∑
λ

∈
S
Q
,
p
(
1
)


exp
(
-
1
2

β


y
~

(
k
)
-
λ

2
)
∑
λ

∈
S
Q
,
p
(
0
)


exp
(
-
1
2

β


y
~

(
k
)
-
λ

2
)
with:
β=η2/V, wherein
η is a finite limit, when the total number K of spreading codes and the length N of the spreading codes tend towards infinity and the quantity α (α=K/N) is less than or equal to 1, of a random variable ηk such that:
y(k)=ηk a(k)+τk,
wherein y(k) is the complex-valued symbol of rank k which corresponds to the transmitted complex symbol a(k) and τk is a random variable representing a noise value produced by multiple access noise and thermal noise filtering,
V is a finite limit of the variance of the random variable τk, β being given by the equation:
β
=
p
k

(
∫
0
1


h

(

2

π


f
)

2



f
)
α

p
_
(
∫
0
1


h

(

2

π


f
)

4



f
∫
0
1


h

(

2

π


f
)

2



f
-
∫
0
1


h

(

2

π


f
)

2



f
)
+
σ
2
wherein
h(e2iπf) is a channel frequency response,
p is a mean power associated with all the K spreading codes,
pk is a power associated with the spreading code of the user of rank k,
σ2 is a thermal noise variance,
{tilde over (y)}(k) is a standardized complex-valued symbol such that {tilde over (y)}(k)=y(k)/η,
λ is a reference complex symbol of QAM constellation associated with the M-QAM modulation,
S(1)I,q is a partition of the complex plane which brings together all the reference complex symbols having a “1” at the position (I, q),
S(0)I,q is a partition of the complex plane which brings together all the reference complex symbols having a “0” at the position (I, q),
S(1)Q,p is a partition of the complex plane which brings together all the reference complex symbols having a “1” at the position (Q, p), and
S(0)Q,p is a partition of the complex plane which brings together all the reference complex symbols having a “0” at the position (Q, p).

286 more elements

A method for soft demapping, by Logarithm of the Likelihood Ratio calculation, of complex-valued symbols (y(1), y(2), . . . , y(k), . . . , y(K)) corresponding to transmitted complex symbols (a(1), a(2), . . . , a(k), . . . , a(K)) contained in a signal transmitted by a code division multiple access multi-carrier transmitter and constituted by binary data coded by convolutional coding and quadrature modulated according to a M-QAM modulation,

wherein M is an integer equal to 2m and m is an integer greater than or equal to 1, a complex symbol transmitted a(k) of rank k, k=1, 2, . . . , K being spread by a spreading code associated with a user of rank k, the binary data coded by convolutional coding constituting, for the user of rank k, a sequence of bits {bI,1;
bI,2;
. . . , bI,q;
. . . , bI,r;
bQ,l;
bQ,2;
. . . , bQ,p;
. . . , bQ,s}, wherein the bit bI,q is the qth coded bit of an in-phase part of the transmitted signal and the bit bQ,p is the pth coded bit of a quadrature part of the transmitted signal, wherein the method includes a Logarithm of the Likelihood Ratio (LogLR) (bI,q) calculation of the bit bI,q and a Logarithm of the Likelihood Ratio (LogLR) (bQ,p) calculation of the bit bQ,p so that:
Log


LR

(
bI
,
q
)
=
ln

∑
λ

∈
S
Q
,
p
(
1
)


exp
(
-
1
2

β


y
~

(
k
)
-
λ

2
)
∑
λ

∈
S
Q
,
p
(
0
)


exp
(
-
1
2

β


y
~

(
k
)
-
λ

2
)
and
Log


LR

(
bQ
,
p
)
=
ln

∑
λ

∈
S
Q
,
p
(
1
)


exp
(
-
1
2

β


y
~

(
k
)
-
λ

2
)
∑
λ

∈
S
Q
,
p
(
0
)


exp
(
-
1
2

β


y
~

(
k
)
-
λ

2
)
with:
β=η2/V, wherein
η is a finite limit, when the total number K of spreading codes and the length N of the spreading codes tend towards infinity and the quantity α (α=K/N) is less than or equal to 1, of a random variable ηk such that:
y(k)=ηk a(k)+τk,
wherein y(k) is the complex-valued symbol of rank k which corresponds to the transmitted complex symbol a(k) and τk is a random variable representing a noise value produced by multiple access noise and thermal noise filtering,
V is a finite limit of the variance of the random variable τk, β being given by the equation:
β
=
p
k
σ
2

(
∫
0
1


h

(

2

π


f
)

2

h

(

2

π


f
)

2
+
σ
2
α

p
_



f
)
∫
0
1


f

h

(

j


2

π


f
)

2
+
σ
2
α

p
_


wherein
h(e2iπf) is a channel frequency response,
p is a mean power associated with all the K spreading codes,
pk is a power associated with the spreading code of the user of rank k,
σ2 is a thermal noise variance,
{tilde over (y)}(k) is a standardized complex-valued symbol such that {tilde over (y)}(k)=y(k)/η,
λ is a reference complex symbol of QAM constellation associated with the M-QAM modulation,
S(1)I,q is a partition of the complex plane which brings together all the reference complex symbols having a “1” at the position (I, q),
S(0)I,q is a partition of the complex plane which brings together all the reference complex symbols having a “0” at the position (I, q),
S(1)Q,p is a partition of the complex plane which brings together all the reference complex symbols having a “1” at the position (Q, p), and
S(0)Q,p is a partition of the complex plane which brings together all the reference complex symbols having a “0” at the position (Q, p).

265 more elements

A method for soft demapping, by Logarithm of Likelihood Ratio calculation, of complex-valued symbols (y(1), y(2), . . . , y(k), . . . , y(K)) corresponding to transmitted complex symbols (a(1), a(2), . . . , a(k), . . . , a(K)) contained in a signal transmitted by a code division multiple access multi-carrier transmitter and constituted by binary data coded by convolutional coding and quadrature modulated according to a M-QAM modulation,

wherein M is an integer equal to 2m and m is an integer greater than or equal to 1, a complex symbol transmitted a(k) of rank k, k=1, 2, . . . , K being spread by a spreading code associated with a user of rank k, the binary data coded by convolutional coding constituting, for the user of rank k, a sequence of bits {bI,1;
bI,2;
. . . , bI,q;
. . . , bI,r;
bQ,I;
bQ,2;
. . . , bQ,p;
. . . ,bQ,s}, wherein the bit bI,q is the qth coded bit of an in-phase part of the transmitted signal and the bit bQ,p is the pth coded bit of a quadrature part of the transmitted signal, wherein the method includes a Logarithm of the Likelihood Ratio (LogLR) (bI,q) calculation of the bit bI,q and a Logarithm of the Likelihood Ratio (LogLR) (bQ,p) calculation of the bit bQ,p so that:
Log


LR

(
bI
,
q
)
=
ln

∑
λ

∈
S
I
,
q
(
1
)


exp
(
-
1
2

β


y
~

(
k
)
-
λ

2
)
∑
λ

∈
S
I
,
q
(
0
)


exp
(
-
1
2

β


y
~

(
k
)
-
λ

2
)
and
Log


LR

(
bQ
,
p
)
=
ln

∑
λ

∈
S
Q
,
p
(
1
)


exp
(
-
1
2

β


y
~

(
k
)
-
λ

2
)
∑
λ

∈
S
Q
,
p
(
0
)


exp
(
-
1
2

β


y
~

(
k
)
-
λ

2
)
with:
β=η2/V, wherein
η is a finite limit, when the total number K of spreading codes and the length N of the spreading codes tend towards infinity and the quantity α (α=K/N) is less than or equal to 1, of a random variable ηk such that:
y(k)=ηk a(k)+τk,
wherein y(k) is the complex-valued symbol of rank k which corresponds to the transmitted complex symbol a(k) and τk is a random variable representing a noise value produced by multiple access noise and thermal noise filtering,
V is a finite limit of the variance of the random variable τk, β being given by the equation:
1
=
∫
0
1

1
σ
2

h

(

j


2

π


f
)

2

A

(
α
,
p
k
,
β
)
+
α
.
B

(
α
,
p
k
,
β
)



f
with

:
A

(
α
,
p
k
,
β
)
=
(
1
-
α
)

β
p
k
+
α
.
β
.

m

(
p
k
,
β
)
B

(
α
,
p
k
,
β
)
=
1
-
p
k
.
m

(
p
k
,
β
)
m

(
p
k
,
β
)
=
∫
-
∞
+
∞

μ
power

(
u
)
p
k
+
β
.
u



u
wherein:
μpower(p) is a limit power distribution in relation to the different powers associated with the different spreading codes, when N and K tend towards infinity, α=K/N remaining constant,
h(e2iπf) is a channel frequency response,
pk is a power associated with the spreading code of the user of rank k,
{tilde over (y)}(k) is a standardized complex-valued symbol such that {tilde over (y)}(k)=y(k)/η,
λ is a reference complex symbol of QAM constellation associated with the M-QAM modulation,
S(1)I,q is a partition of the complex plane which brings together all the reference complex symbols having a “1” at the position (I, q),
S(0)I,q is a partition of the complex plane which brings together all the reference complex symbols having a “0” at the position (I, q),
S(1)Q,p is a partition of the complex plane which brings together all the reference complex symbols having a “1” at the position (Q, p), and
S(0)Q,p is a partition of the complex plane which brings together all the reference complex symbols having a “0” at the position (Q, p).

326 more elements

Dependencies are inferred from the claim text.

Method for the Flexible Demodulation of Estimated Complex Symbols

Life of this patent

Method for the Flexible Demodulation of Estimated Complex Symbols

Life of this patent

Abstract

Claims19