Granted Patent B2

US 12,556,946 · App. 18/633,392 · Granted Feb 17, 2026

Inventors: Maged F. Barsoum (San Jose, CA); Christopher R. Jones (Pacific Palisades, CA)

Assignee: Constellation Designs, LLC

H04W24/02H03M13/255H03M13/6325H04B15/00H04B17/336H04L1/0003H04L1/0009H04L27/3405H04L27/3483H04L27/3809

Abstract

Claims (29)

1 . A method of selecting points of a multidimensional geometrically shaped symbol constellation used in a communication system, comprising:

selecting an appropriate constellation size and a desired parallel decode capacity per dimension for a multidimensional geometrically shaped symbol constellation;

estimating an initial SNR at which a communication system is likely to operate; and

iteratively optimizing labels and locations of points of the multidimensional geometrically shaped symbol constellation to maximize parallel decode capacity until a predetermined improvement in Signal-to-Noise Ratio (SNR) performance of the multidimensional geometrically shaped symbol constellation relative to a multidimensional symbol constellation that maximizes d min has been achieved, where d min is a minimum distance between locations of constellation points.

2 . The method of claim 1 , wherein the iterative optimization iterates until a predetermined SNR gain is achieved.

3 . The method of claim 2 , wherein the predetermined SNR gain is 0.5 db.

4 . The method of claim 1 , further comprising:

selecting an SNR performance of each optimized multidimensional geometrically shaped symbol constellation as a new SNR at which to optimize the labels and locations of the points of the multidimensional geometrically shaped symbol constellation based on the parallel decode capacity in each iteration; and

iterating until an SNR performance obtained by the optimized multidimensional geometrically shaped symbol constellation converges to within a predetermined amount of the SNR performance of the optimized multidimensional geometrically shaped symbol constellation obtained in a previous iteration.

5 . The method of claim 1 , wherein the initial SNR estimate is an SNR required by a multidimensional constellation that maximizes d min to obtained the desired capacity.

6 . The method of claim 1 , wherein the iterative optimization is an iterative numerical process that selects the labels and locations of the multidimensional geometrically shaped symbol constellation with a highest measure of parallel decode capacity.

7 . The method of claim 6 , wherein the iterative optimization is subject to additional constraints.

8 . The method of claim 7 , wherein the additional constraints include at least one constraint selected from a group consisting of: a constant energy constraint, a peak to average power constraint, and a labeling constraint.

9 . The method of claim 1 , wherein the iterative optimization yields a multidimensional geometrically shaped symbol constellation having a suboptimal capacity.

10 . The method of claim 1 , wherein the points of the optimized multidimensional geometrically shaped symbol constellation are modified using at least one process selected from a group consisting of: approximation, quantization and slight variation.

11 . A method of selecting points of a multidimensional geometrically shaped symbol constellation used in a communication system, comprising:

selecting an appropriate constellation size and a desired parallel decode capacity per dimension for a multidimensional geometrically shaped symbol constellation;

selecting an initial Signal-to-Noise Ratio (SNR) estimate as an SNR required by a multidimensional constellation that maximizes d min to obtained the desired capacity, where d min is a minimum distance between locations of constellation points; and

iteratively optimizing labels and locations of the points of the multidimensional geometrically shaped symbol constellation to maximize parallel decode capacity to obtain an improvement in SNR performance of the multidimensional geometrically shaped constellation relative to a multidimensional constellation that maximizes d min by:

optimizing the labels and locations of the points of the multidimensional geometrically shaped symbol constellation at the initial SNR based on the parallel decode capacity to obtain an initial optimized multidimensional geometrically shaped symbol constellation;

for each subsequent iteration, selecting an SNR performance of each optimized multidimensional geometrically shaped symbol constellation from a previous iteration as a new SNR at which to optimize the labels and locations of the points of the multidimensional geometrically shaped symbol constellation based on the parallel decode capacity; and

iterating until an SNR performance obtained by the optimized multidimensional geometrically shaped symbol constellation converges to within a predetermined amount of an SNR performance of the optimized multidimensional geometrically shaped symbol constellation obtained in a previous iteration.

12 . The method of claim 11 , wherein the iterative optimization iterates until a predetermined SNR gain is achieved.

13 . The method of claim 12 , wherein the predetermined SNR gain is 0.5 db.

14 . The method of claim 11 , wherein the iterative optimization is an iterative numerical process that selects the multidimensional geometric symbol constellation point labels and locations with a highest measure of parallel decode capacity.

15 . The method of claim 14 , wherein the iterative optimization is subject to additional constraints.

16 . The method of claim 15 , wherein the additional constraints include at least one constraint selected from the group consisting of: a constant energy constraint, a peak to average power constraint, and a labeling constraint.

17 . The method of claim 11 , wherein the iterative optimization yields a multidimensional geometrically shaped symbol constellation having a suboptimal capacity.

18 . The method of claim 11 , wherein the points of the optimized multidimensional geometrically shaped constellation are modified using at least one process selected from a group consisting of: approximation, quantization and slight variation.

Assignments (2)

Continuity (11)

References Cited (400)