Uplink Soft-Value Calculation With Reduced MLD Complexity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current wireless communication systems face a trade-off between complexity and performance in interference cancellation, with advanced detection schemes like MLD having high computational complexity, and existing methods not effectively reducing complexity while maintaining accuracy.
Innovation Solution
A receiver node that evaluates bit sequences by discarding certain possibilities and calculating metrics to determine the most likely bit sequence, with the option to add metric data and insert missing bits to ensure accurate estimation, using techniques like M-algorithm and soft value calculations based on Euclidean distances.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If maximum likelihood detection (MLD) is used for interference cancellation, then performance in terms of accuracy is improved, but computational complexity increases exponentially
Solution Approach 1:
The patent segments the bit sequences into two sets: those that are discarded (eliminated) and those that are kept (processed). This segmentation allows the receiver to focus computational resources only on the most likely bit sequences, reducing complexity while maintaining accuracy. The M-algorithm further segments the search space by keeping only the M most promising candidates at each stage.
Solution Approach 2:
Instead of processing all possible bit sequences (excessive action), the patent applies partial action by discarding unlikely sequences and keeping only the most probable ones. This partial processing approach reduces computational complexity while maintaining sufficient accuracy for reliable data recovery.
2Measurement precision
If all possible bit sequences are evaluated to ensure accuracy, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent applies local quality by treating different bit sequences differently based on their likelihood. High-likelihood sequences are kept and processed in detail, while low-likelihood sequences are discarded. This differentiated processing approach maintains accuracy for the most important cases while reducing overall complexity.
Solution Approach 2:
The patent changes the parameter of processing depth dynamically. For the most likely bit sequences, deep processing is applied (keeping and evaluating metrics), while for less likely sequences, processing is terminated early (discarding). This parameter change strategy optimizes the balance between accuracy and complexity.
3Device complexity
If simplified detection schemes are used to reduce complexity, then device complexity is reduced, but performance in terms of accuracy deteriorates
Solution Approach 1:
The patent performs preliminary action by calculating metrics for bit sequences before final decision-making. The metrics (such as Euclidean distance) are computed in advance to guide the discarding and keeping decisions, enabling simplified yet accurate detection. This preliminary computation allows the receiver to make informed decisions about which sequences to process further.
4Measurement precision
If metric data is added for incomplete signal streams, then soft value calculations are improved, but processing time increases
Solution Approach 1:
The patent uses copying by creating an added bit sequence that replicates the structure of complete sequences. This copied sequence is then used to compute metrics for incomplete signal streams, enabling soft value calculations without requiring full sequence processing. The copying approach improves calculation quality while controlling processing time.
Data Source
AI summary
A node comprising a receiver for i) receiving at least two signals streams comprising bit sequences and ii) evaluating which bit sequence that is most likely to have been received for a certain sent symbol for each signal stream. The receiver for calculating metrics indicative of which bit sequence that initially is most likely to correspond to a certain sent symbol, the metrics being used a soft value calculation where the receiver is arranged for addition of metric data for a certain signal stream corresponding to an added bit sequence for each case where the available metrics are incomplete for performing the estimation. The missing bit is inserted in the added bit sequence and chosen such that it corresponds to a symbol with the shortest Euclidian distance to the symbol with the said corresponding bit sequence initially being indicated as most likely to correspond to a certain sent symbol.