Reduced Reference Sequence Decoding for Control Bit Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In wireless communication systems, the existing methods for decoding control bits are complex and require significant memory and computational resources due to the large reference sequence space, which is exacerbated by the need for error robustness and limited bandwidth.
Innovation Solution
A method and system that reduce the reference sequence space by determining candidate sequences from a reduced reference sequence space, calculating correlation energies, and selecting complete candidate sequences to determine the sequence of bits, thereby reducing memory size and computational complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a large reference sequence space is used for maximum-likelihood sequence estimation, then decoding reliability is improved, but memory size and computational complexity increase exponentially
Solution Approach 1:
The patent segments the reference sequence space by identifying and removing redundant sequences. Instead of processing all 2^N possible sequences, the method divides the sequence space into equivalence classes where sequences that differ only in non-informative bits are grouped together. This segmentation reduces the effective search space from exponential to polynomial complexity while preserving all unique decoding information.
Solution Approach 2:
The patent extracts and removes redundant sequences from the reference sequence space. By identifying sequences that provide no additional decoding information (those differing only in bits not involved in the encoding mapping), the method extracts the essential unique sequences and discards the rest. This extraction process reduces memory requirements and computational load while maintaining decoding reliability.
2Reliability
If forward error correction codes with high redundancy are used, then transmission reliability is improved, but the number of control bits that can be transmitted decreases due to limited bandwidth
Solution Approach 1:
The patent applies partial action by computing correlation energies for only a subset of sequences (the non-redundant ones) rather than all possible sequences. This partial processing approach maintains sufficient decoding accuracy while significantly reducing the computational resources required, effectively allowing more control bits to be transmitted within the same bandwidth constraints.
3Measurement precision
If conventional maximum-likelihood sequence estimation is used, then optimal decoding is achieved, but processing time increases due to exponential complexity
Solution Approach 1:
The patent segments the sequence space to identify unique equivalence classes, reducing the number of sequences that need to be evaluated. By processing only the essential unique sequences rather than all possible sequences, the method maintains optimal decoding accuracy while reducing processing time from exponential to polynomial complexity.
Solution Approach 2:
The patent creates a compressed representation (copy) of the sequence space that contains only the unique decoding information. Instead of working with the full exponential sequence space, the method creates and processes a condensed version that preserves all necessary decoding relationships, significantly accelerating processing while maintaining accuracy.
Data Source
AI summary
A system and method is provided for estimating a sequence of N bits ({circumflex over (x)}0{circumflex over (x)}1 . . . {circumflex over (x)}N−1) corresponding to a received sequence of M digital data (r0r1 . . . rM−1). The method includes determining candidate sequences of MRS digital data from a reduced reference sequence space comprising 2N<sub2>RS </sub2>reduced reference sequences of MRS reference digital data (s0s1 . . . sM<sub2>RS</sub2>−1), MRS being less than M, and 2N<sub2>RS </sub2>being less than or equal to 2N. The method further includes making up each candidate sequence with remaining reference symbols to obtain at least one complete candidate sequence of M digital data, and determining the sequence of N bits ({circumflex over (x)}0{circumflex over (x)}1 . . . {circumflex over (x)}N−1) from the complete candidate sequences.