Weighted Sequential Decoding With Sub-Block Metrics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing decoding techniques for wireless communication systems face challenges in efficiently decoding signals due to high computational complexity, especially with increasing constellation sizes and the number of antennas, leading to suboptimal performance and error rates.
Innovation Solution
A decoder is introduced that divides the received signal vector into sub-vectors using a channel matrix decomposition, calculates sub-block weighting coefficients, and applies a sequential decoding algorithm with modified metrics to reduce complexity and improve performance, allowing for flexible bias adjustments based on signal-to-noise ratio and tree level parameters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If exhaustive search ML decoding is used, then decoding performance is optimal, but computational complexity increases with constellation size and number of antennas
Solution Approach 1:
The patent divides the received signal vector into multiple sub-vectors and processes them through parallel decoding paths. Each sub-vector is decoded independently using simplified metrics, avoiding the need for exhaustive search while maintaining acceptable performance. This segmentation reduces computational complexity from exponential to polynomial in the constellation size.
Solution Approach 2:
The patent applies different decoding strategies to different parts of the signal. Specifically, it uses simplified local metrics for individual sub-vectors rather than a global exhaustive search, allowing faster local decisions that collectively approximate the optimal solution without the full computational burden.
2Device complexity
If suboptimal low-complexity decoders (ZF, ZF-DFE, MMSE) are used, then computational complexity is reduced, but decoding performance and error performance deteriorate
Solution Approach 1:
The patent introduces an intermediary sequential decoding structure that bridges the gap between simple linear decoders and exhaustive search. It uses a stack-based approach with modified metrics that incorporate interference cancellation, providing intermediate performance levels that are better than ZF/MMSE but computationally feasible.
Solution Approach 2:
The patent employs dynamic metric adjustment in the sequential decoding process. The metrics are updated based on previously decoded symbols and current channel conditions, allowing the decoder to adapt its search strategy in real-time to maintain performance while controlling complexity.
3Device complexity
If sequential decoders with global stack are used, then complexity is reduced compared to exhaustive search, but performance degrades with increasing constellation size and antennas
Solution Approach 1:
The patent segments the global decoding problem into multiple sub-problems, each handled by a separate decoding path with its own stack. This segmentation prevents any single stack from becoming too large and complex, while the combination of parallel paths maintains overall performance.
Solution Approach 2:
The patent modifies the decoding metrics by incorporating weighting factors and interference cancellation terms. These parameter changes to the metric calculation allow the sequential decoder to achieve better performance without increasing the fundamental complexity of the stack-based search approach.
Data Source
AI summary
Embodiments of the invention provide a decoder (10) for decoding a signal received through a transmission channel in a communication system, the signal carrying information symbols selected from a given set of values and being associated with a signal vector, the transmission channel being represented by a channel matrix. The decoder comprises:a sub-block division unit (12) configured to divide the received signal vector into a set of sub-vectors in correspondence with a division of a matrix related to said channel matrix;at least one weighting coefficient calculation unit (14) configured to calculate a sub-block weighting coefficient for each sub-vector,at least one symbol estimation unit (11) for recursively determining estimated symbols representative of the transmitted symbols carried by the data signal from information stored in a stack.The at least one symbol estimation unit is configured to apply at least one iteration of a sequential decoding algorithm, the sequential decoding algorithm comprising iteratively filling a stack by expanding child nodes of a selected node of a decoding tree comprising a plurality of nodes, each node of the decoding tree corresponding to a candidate component of a symbol of at least a part of the received signal and each node being assigned an initial metric. The symbol estimation unit is further configured to calculate a modified metric for at least one node of the expanded child nodes from the metric associated with the at least one node and from the sub-block weighting coefficient calculated for the sub-vector to which the at least one node belongs, symbol estimation unit being configured to assign the modified metric to the at least one node.