Convolutional Code Puncturing Patterns for Feasible Error-Rate Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Identifying puncturing patterns for punctured convolutional codes that exhibit good error correcting performance is not straightforward, especially for longer pattern lengths, making exhaustive searches computationally infeasible.
Innovation Solution
A method for generating puncturing patterns by simulating data transmission over a noisy channel and determining transmission error rates for candidate patterns, excluding those that puncture more bits from the generator polynomial with fewer non-zero coefficients, thereby reducing the number of patterns to test and improving efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If exhaustive search is used to identify puncturing patterns, then the error-correcting performance can be optimized, but the computational complexity becomes infeasible for longer pattern lengths
Solution Approach 1:
The patent segments the search space of puncturing patterns by imposing a constraint: the number of punctured bits from the first generator polynomial must be greater than or equal to the number of punctured bits from the second generator polynomial. This segmentation divides the exhaustive search space into manageable subsets, allowing systematic evaluation of candidate patterns without requiring complete exhaustive search, thus reducing computational complexity while maintaining error-correcting performance optimization.
Solution Approach 2:
The patent changes the search parameters by introducing a specific constraint on the distribution of punctured bits between different generator polynomials. Instead of searching all possible puncturing patterns uniformly, the method focuses on patterns where the puncturing distribution satisfies the specified inequality constraint, thereby changing the search parameters to achieve feasible computational complexity while preserving the ability to identify high-performance patterns.
2Productivity
If the number of candidate puncturing patterns is reduced by applying constraints, then the computational burden is reduced, but the risk of missing optimal patterns increases
Solution Approach 1:
The patent applies preliminary action by pre-establishing the constraint condition before the actual search process. By determining in advance that candidate patterns must satisfy the puncturing distribution inequality, the method filters the search space beforehand. This preliminary filtering reduces the number of candidates requiring full simulation and evaluation, improving productivity while the systematic nature of the constraint ensures that optimal patterns within the constrained space are not missed.
Solution Approach 2:
The patent incorporates feedback mechanisms by evaluating candidate patterns through simulation and measuring their error rates. The feedback from simulation results allows the method to identify high-performance patterns within the constrained search space. The process uses feedback to confirm that the constraint-based approach successfully identifies patterns with good error-correcting performance, validating that productivity improvement does not compromise reliability.
Data Source
AI summary
A method of generating a puncturing pattern for use with a 1/n convolutional code having a predetermined code rate is disclosed. The 1/n convolutional code uses a predetermined set of binary generator polynomials comprising a first generator polynomial and a second generator polynomial, wherein the second generator polynomial has more non-zero coefficients than the first generator polynomial. For each of a plurality of candidate puncturing patterns, transmission of data over a noisy channel is simulated, the data being encoded using the 1/n convolutional code punctured in accordance with the respective puncturing pattern. A transmission error rate is determined for each candidate, and the error rates are compared to identify a candidate that has a lowest error rate. The plurality of candidate puncturing patterns includes only patterns that puncture no more coded bits generated by the second generator polynomial than they puncture coded bits generated by the first generator polynomial.