Sequential Decoder Metrics With Tree-Level Weighting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing wireless communication systems face challenges in decoding received signals due to high complexity and poor performance in spectral efficiency and error rates, especially with increasing constellation sizes and antenna numbers, as current decoding methods require high computational complexity and lack flexibility in performance-complexity tradeoffs.
Innovation Solution
The introduction of a tree-level dependent weighting coefficient for sequential decoders, such as Stack, Fano, or SB-Stack decoders, which adjusts the metric calculation based on Signal-to-Noise Ratio (SNR), channel gains, and tree levels, optimizing the performance-complexity tradeoff by varying the weighting coefficient across different levels of the decoding tree.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If exhaustive search ML decoding is used, then optimal decoding performance is achieved, but computational complexity increases significantly with constellation size and antenna number
Solution Approach 1:
The patent segments the exhaustive search space by introducing a tree structure that divides the decoding problem into hierarchical levels. Each level processes a subset of constellation points, allowing the decoder to explore promising paths while pruning unlikely candidates, thus reducing overall computational complexity while maintaining near-optimal performance
Solution Approach 2:
The patent applies local quality by using different metric calculations and search strategies for different regions of the decoding tree. High-probability regions receive more thorough exploration with refined metrics, while low-probability regions are pruned earlier, optimizing the trade-off between performance and complexity in different local areas of the search space
2Device complexity
If suboptimal low-complexity decoders like ZF or MMSE are used, then computational complexity is reduced, but error performance and diversity order deteriorate
Solution Approach 1:
The patent introduces dynamic adaptation by making the decoding strategy flexible and adjustable based on channel conditions, constellation size, and performance requirements. The decoder can dynamically switch between different search depths, metric types, and pruning thresholds, allowing it to achieve good error performance when needed while maintaining low complexity when channel conditions permit
Solution Approach 2:
The patent changes key parameters such as the metric calculation formula, search threshold values, and tree expansion depth based on channel signal-to-noise ratio and constellation characteristics. By adapting these parameters to current conditions, the decoder achieves better error performance than fixed low-complexity methods without incurring the full complexity of exhaustive search
3Device complexity
If fixed bias sequential decoding is used, then computational complexity is reduced compared to exhaustive search, but flexibility in performance-complexity tradeoff is limited
Solution Approach 1:
The patent transforms the static fixed-bias approach into a dynamic system where the bias parameter and other decoding parameters are adjusted in real-time based on channel conditions, desired performance targets, and observed decoding progress. This dynamic adaptation provides continuous flexibility in the performance-complexity tradeoff rather than fixed operating points
Solution Approach 2:
The patent incorporates feedback mechanisms where the decoder monitors its own performance metrics and decoding progress, then uses this information to adjust subsequent decoding parameters. This closed-loop approach allows the system to automatically optimize the performance-complexity tradeoff based on actual channel conditions and performance requirements
Data Source
Figure 1
Figure 2
Figure 3
AI summary
There is provided a decoder for sequentially decoding a data signal received through a transmission channel in a communication system, said data signal carrying transmitted symbols, said decoder comprising a symbol estimation unit (301) configured to determine estimated symbols representative of the transmitted symbols carried by the received signal from information stored in a stack, said symbol estimation unit (301) being configured to iteratively fill the 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 said data signal and each node being assigned a metric, the stack being filled at each iteration with a set of expanded child nodes and being ordered by increasing values of the metrics assigned to the nodes, the selected node for each iteration corresponding to the node being assigned the lowest metric in the stack, the decoder comprising a metric determination unit (302) configured to determine an initial metric for each child node of said set of expanded child nodes, wherein the decoder further comprises a modified metric calculation unit (303) configured to calculate a modified metric for at least one of the expanded child nodes from the metric associated with said expanded child node and a weighting coefficient, said weighting coefficient being a function of the level of said node in the decoding tree, the decoder assigning said modified metric to said at least one of the expanded child nodes.