Soft-Decision Demodulation Using Correlator Phase Statistics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing communication systems lack a mathematically rigorous and robust method for converting correlator output magnitudes into likelihoods of orthogonal communication sequences and bit values, leading to inaccurate soft-decision demodulation under unfavorable channel conditions.
Innovation Solution
A communication unit and method for soft-decision demodulation using a receiver with a bank of correlators, employing a demodulator and de-mapper circuit to determine statistics from aggregated correlator output distributions, enabling high-quality soft-decisions through parameterized distribution representations and likelihood calculations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If hard decision demodulation is used with a bank of correlators, then the system is simple to implement, but error correction capability deteriorates under unfavorable channel conditions
Solution Approach 1:
The patent transforms the hard decision output (binary 0 or 1) into soft decision outputs by calculating logarithmic likelihood ratios (LLRs) for each bit. This parameter transformation enables the system to provide reliability information about each bit decision, allowing the channel decoder to perform more effective error correction while maintaining implementation feasibility through the use of correlator outputs.
2Reliability
If channel decoding is applied to correct bit errors, then error correction improves, but the decoder's ability is limited when operating on hard decisions
Solution Approach 1:
The patent introduces an intermediary processing stage between the correlator and channel decoder that converts hard decisions into soft decisions using likelihood calculations. This intermediary layer provides the channel decoder with rich probabilistic information about each bit, significantly enhancing error correction capability without adding complex decoder architecture.
3Reliability
If soft-decision demodulation is implemented, then error correction capability improves, but the conversion method from correlator outputs to likelihoods becomes complex
Solution Approach 1:
The patent replaces complex probabilistic calculations with a simplified mathematical transformation using logarithmic likelihood ratios. Instead of computing full probability distributions, the system uses the correlator output magnitudes directly in a logarithmic transformation, which maintains mathematical rigor while dramatically simplifying the conversion process and making it suitable for practical implementation.
4Measurement precision
If all correlator output distributions are used for soft-decision calculation, then accuracy improves, but computational complexity and sensitivity to channel variations increase
Solution Approach 1:
The patent extracts only the essential information needed for soft-decision demodulation from the correlator outputs - specifically the magnitudes of the correlator outputs. By focusing on these key parameters and using them to calculate LLRs, the system achieves high accuracy without the need to process all possible distribution parameters, thereby reducing computational complexity and improving robustness to channel variations.
Data Source
AI summary
A communication unit for performing soft-decision demodulation comprises a receiver that receives a transmitted signal having a first set of bits comprising k bits, selected from a set of 2k possible signals according to values of the k bits, and a second set of bits comprising Qm bits based on a phase rotation of the transmitted signal selected from a set of 2Qm possible rotations. The receiver comprises: a demodulator comprising a bank of 2k correlators and is configured to: detect a transmission of each possible transmitted signal, and output 2k phases of the correlator outputs as a third set of inputs. A de-mapper circuit receives the third set of inputs: determines statistics derived from a number of aggregated correlator output phase distributions of the third set of inputs; and calculates and outputs a second set of aposteriori soft bits comprising Qm soft bits.