Ai/ml empowered high order modulation
AssigneeSamsung Electronics Co., Ltd.
InventorsPranav Madadi, Joonyoung Cho, Jianzhong Zhang
Life of this patent
Estimating the term…
Abstract
A two-dimensional constellation for data signals having improved bitwise mutual information of data points is based on a signal-to-noise ratio (SNR) and a code rate where, based on the SNR, data bits are mapped to pre-defined in-phase and quadrature values. The in-phase and quadrature values denote points in the two-dimensional space such that the efficiency of bitwise mutual information is adapted based on the SNR. The mapping is preferably subject to a constraint selected from one of quadrant symmetry Lagrangian (QSL), quadrant symmetry constraint (QSC), or rectangular structure constraint (RSC).
Claims18
A method, comprising:
- receiving configuration information including a signal-to-noise ratio (SNR) and a code rate;
- and
- mapping, based on the SNR, data bits to pre-defined in-phase and quadrature values,
- wherein the in-phase and quadrature values denote points on a two-dimensional (2D)) space such that optimality of bitwise mutual information is adapted based on the SNR.
An apparatus, comprising:
- a transceiver configured to receive configuration information including a signal-to-noise ratio (SNR) and a code rate;
- and
- a controller configured to map, based on the SNR, data bits to pre-defined in-phase and quadrature values,
- wherein the in-phase and quadrature values denote points on a two-dimensional (2D) space such that optimality of bitwise mutual information is adapted based on the SNR.
A method, comprising:
- determining configuration information including a signal-to-noise ratio (SNR) and a code rate for use in mapping, based on the SNR, data bits to pre-defined in-phase and quadrature values;
- and
- transmitting the configuration information,
- wherein the in-phase and quadrature values denote points on a two-dimensional (2D)) space such that optimality of bitwise mutual information is adapted based on the SNR.
Dependencies are inferred from the claim text.