Probabilistic Signal Point Shaping for BICM Capacity

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing probabilistic signal point shaping devices and methods fail to optimize Bit-Interleaved Coded-Modulation (BICM) capacity effectively, particularly in achieving maximum link throughput over a given transmit bandwidth and signal-to-noise ratio.

Innovation Solution

The development of a probabilistic signal point shaping device and method that enhances probabilistic amplitude shaping (PAS) by generating signal point constellations with non-uniform probability distributions, allowing for simultaneous IQ distribution matching and supporting signal point constellations with odd numbers of bits, thereby optimizing BICM capacity.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If uniform signal point constellations are used, then the device complexity is low, but the BICM capacity is not optimized

Engineering Contradiction:
ImproveBICM capacityVSAvoidsignal point constellation complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent changes the probability distribution parameter of signal points from uniform to non-uniform distribution. By assigning different probabilities to different signal points in the constellation, the system optimizes BICM capacity without fundamentally changing the constellation structure, thus achieving capacity improvement with controlled complexity increase

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent applies different probability weights to different signal points locally within the constellation. Instead of treating all signal points uniformly, each signal point is assigned a specific probability based on its position and channel characteristics, allowing local optimization of the transmission efficiency

Inventive Principle:
Principle #3Local quality

2Productivity

If probabilistic amplitude shaping is used, then the channel capacity increases, but the adaptability to different channel types is limited

Engineering Contradiction:
Improvechannel capacityVSAvoidchannel type adaptability
Core Design Contradiction:
ProductivityVSAdaptability or versatility

Solution Approach 1:

The patent extends the probabilistic shaping framework to work with multiple channel types (AWGN, fading channels, etc.) and different modulation schemes. The same probabilistic signal point shaping device can be adapted to various channel conditions by adjusting the probability distribution parameters, making it a universal solution for capacity optimization across different communication scenarios

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Productivity

If signal point constellations with odd numbers of bits are used, then the bandwidth efficiency improves, but the implementation complexity increases

Engineering Contradiction:
Improvebandwidth efficiencyVSAvoidmapping complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent employs dynamic probability assignment for signal points in constellations with odd numbers of bits. Rather than using fixed mapping rules, the system dynamically adjusts the probability distribution to accommodate non-standard constellation sizes, enabling flexible bandwidth-efficient transmissions while managing implementation complexity through adaptive algorithms

Inventive Principle:
Principle #15Dynamics

Data Source

PatentEP3673594B1Probabilistic signal point shaping device and method
Publication Date: 2025.04.09 SONY GROUP CORP
  • EP3673594B1 patent drawingFigure 1~2
  • EP3673594B1 patent drawingFigure 3~5
  • EP3673594B1 patent drawingFigure 6

AI summary

A probabilistic signal point shaping device comprises circuitry configured to map data blocks of input bits of an input bit stream onto mapping symbols, wherein said mapping symbols are distributed according to a predetermined probability distribution and represented by complex-valued signal points or non-uniformly spaced amplitude levels, assign bit labels to said mapping symbols, determine redundancy bits from the bits of said bit labels, transform said redundancy bits into a redundancy rule, and apply the redundancy rule to the mapping symbols to obtain output symbols.