Low complexity distance metrics for maximum likelihood receivers
First Claim
1. A method of determining a data value for a bit of interest in a received data signal, the method comprising:
- selecting a closest 0-bit symbol candidate, at the bit of interest, from a plurality of 0-bit symbol candidates using a symbol selection distance approximation;
selecting a closest 1-bit symbol candidate, at the bit of interest, from a plurality of 1-bit symbol candidates using the symbol selection distance approximation;
calculating a 0-bit distance and a 1-bit distance, wherein the 0-bit distance represents the distance between the received data signal and the selected closest 0-bit symbol candidate, and the 1-bit distance represents the distance between the received data signal and the selected closest 1-bit symbol candidate;
calculating a log likelihood ratio for the bit of interest using i) the calculated 0-bit distance and ii) the calculated 1-bit distance; and
determining the data value for the bit of interest based on the calculated log likelihood ratio;
wherein locations of each of the received data signal, the plurality of 0-bit symbol candidates, and the plurality of 1-bit symbol candidates are identified by coordinates;
wherein the symbol selection distance approximation, as applied to selecting the closest 0-bit symbol candidate, includesestimating first distances between the received data signal and the plurality of 0-bit symbol candidates using norm approximations of vectors from the coordinates of the received data signal to coordinates of the plurality of 0-bit symbol candidates,wherein the symbol selection distance approximation, as applied to selecting the closest 1-bit symbol candidate, includesestimating second distances between the received data signal and the plurality of 1-bit symbol candidates using norm approximations of vectors from the coordinates of the received data signal to coordinates of the plurality of 1-bit symbol candidates,wherein the vector between the coordinates of the received data signal and the coordinates of a 0-bit or 1-bit symbol candidate is represented as zk=zkI+jzkQ,wherein the norm approximation of ∥
zk∥
estimates ∥
zk∥
=max{|zkI|,|zkQ|}+α
·
min{|zkI|,|zkQ|}, where α
is 5/16, andwherein the first or second distances are calculated according to a non-iterative distance approximation according to;
∥
z∥
≈
G(z),where
2 Assignments
0 Petitions
Accused Products
Abstract
Systems and methods are provided for determining a data value for a bit of interest in a received data signal. A closest 0-bit symbol candidate is selected from a plurality of 0-bit symbol candidates using a symbol selection distance approximation. A closest 1-bit symbol candidate is selected from a plurality of 1-bit symbol candidates using the symbol selection distance approximation. A 0-bit distance between the received data signal and the selected 0-bit symbol candidate is determined, and a 1-bit distance between the received data signal and the selected 1-bit symbol candidate is determined. A log likelihood ratio is calculated for the bit of interest using the determined 0-bit distance and the determined 1-bit distance.
14 Citations
28 Claims
-
1. A method of determining a data value for a bit of interest in a received data signal, the method comprising:
-
selecting a closest 0-bit symbol candidate, at the bit of interest, from a plurality of 0-bit symbol candidates using a symbol selection distance approximation; selecting a closest 1-bit symbol candidate, at the bit of interest, from a plurality of 1-bit symbol candidates using the symbol selection distance approximation; calculating a 0-bit distance and a 1-bit distance, wherein the 0-bit distance represents the distance between the received data signal and the selected closest 0-bit symbol candidate, and the 1-bit distance represents the distance between the received data signal and the selected closest 1-bit symbol candidate; calculating a log likelihood ratio for the bit of interest using i) the calculated 0-bit distance and ii) the calculated 1-bit distance; and determining the data value for the bit of interest based on the calculated log likelihood ratio; wherein locations of each of the received data signal, the plurality of 0-bit symbol candidates, and the plurality of 1-bit symbol candidates are identified by coordinates; wherein the symbol selection distance approximation, as applied to selecting the closest 0-bit symbol candidate, includes estimating first distances between the received data signal and the plurality of 0-bit symbol candidates using norm approximations of vectors from the coordinates of the received data signal to coordinates of the plurality of 0-bit symbol candidates, wherein the symbol selection distance approximation, as applied to selecting the closest 1-bit symbol candidate, includes estimating second distances between the received data signal and the plurality of 1-bit symbol candidates using norm approximations of vectors from the coordinates of the received data signal to coordinates of the plurality of 1-bit symbol candidates, wherein the vector between the coordinates of the received data signal and the coordinates of a 0-bit or 1-bit symbol candidate is represented as zk=zkI+jzkQ, wherein the norm approximation of ∥
zk∥
estimates ∥
zk∥
=max{|zkI|,|zkQ|}+α
·
min{|zkI|,|zkQ|}, where α
is 5/16, andwherein the first or second distances are calculated according to a non-iterative distance approximation according to;
∥
z∥
≈
G(z),where - View Dependent Claims (3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15)
-
-
2. A method of determining a data value for a bit of interest in a received data signal, the method comprising:
-
selecting a closest 0-bit symbol candidate, at the bit of interest, from a plurality of 0-bit symbol candidates using a symbol selection distance approximation; selecting a closest 1-bit symbol candidate, at the bit of interest, from a plurality of 1-bit symbol candidates using the symbol selection distance approximation; calculating a 0-bit distance and a 1-bit distance, wherein the 0-bit distance represents the distance between the received data signal and the selected closest 0-bit symbol candidate, and the 1-bit distance represents the distance between the received data signal and the selected closest 1-bit symbol candidate; calculating a log likelihood ratio for the bit of interest using i) the calculated 0-bit distance and ii) the calculated 1-bit distance; and determining the data value for the bit of interest based on the calculated log likelihood ratio; wherein locations of each of the received data signal, the plurality of 0-bit symbol candidates, and the plurality of 1-bit symbol candidates are identified by coordinates; wherein the symbol selection distance approximation, as applied to selecting the closest 0-bit symbol candidate, includes estimating first distances between the received data signal and the plurality of 0-bit symbol candidates using norm approximations of vectors from the coordinates of the received data signal to coordinates of the plurality of 0-bit symbol candidates, wherein the symbol selection distance approximation, as applied to selecting the closest 1-bit symbol candidate, includes estimating second distances between the received data signal and the plurality of 1-bit symbol candidates using norm approximations of vectors from the coordinates of the received data signal to coordinates of the plurality of 1-bit symbol candidates, wherein the vector between the coordinates of the received data signal and the coordinates of a 0-bit or 1-bit symbol candidate is represented as zk=zkI+jzkQ, wherein the norm approximation of ∥
zk∥
estimates ∥
zk∥
=max{|zkI|,|zkQ|}+α
·
min{∥
zkI|,|zkQ|}, where α
is 5/16, andwherein the first or second distances are calculated according to an iterative distance approximation according to;
∥
z∥
≈
{hacek over (G)}(z)(N),where
{hacek over (G)}(z)(1)=g(z1),where
{hacek over (G)}(z)(k)=max{g(zk),G(z)(k−
1)}+α
·
min{g(zk),G(z)(k−
1)}, k=2 . . . N,where
g(zk)=max{|zkI|,|zkQ|}+α
·
min{|zkI|,|zkQ|}, k=1 . . . N.- View Dependent Claims (16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28)
-
25. The method of claim 24, wherein the log likelihood ratio is calculated according to:
-
L(bk(i))≈
1/σ
2{||y−
Hx0||2−
||y−
Hx1||2},where ∥
y−
Hx0∥
is the 0-bit distance, |y−
Hx1∥
is the 1-bit distance, L(bk(i)) is the log likelihood ratio for a bit of interest, and σ
2 is a variance of noise.
-
-
26. The method of claim 25, wherein the bit of interest is determined to be a 0 or a 1 based on the log likelihood ratio.
-
27. The method of claim 2, wherein the received data signal is a multiple input, multiple output (MIMO) signal.
-
28. The method of claim 2, wherein the received data signal is a single carrier signal.
-
Specification