Adaptive SOMA MIMO Demapper for OFDM Complexity Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing receiver structures for MIMO/OFDM/BICM systems face high computational complexity, particularly with MAP-based and MaxLogMAP-based decoders, which become intractable as the number of transmit antennas and QAM symbols increase, leading to suboptimal bit-error-rate performance and increased complexity.
Innovation Solution
The implementation of a tree position adaptive Soft Output M-algorithm (SOMA) receiver structure that reduces complexity by intelligently searching a detection tree based on channel state information and extrinsic information, using a fraction of candidates and adapting parameters like M, T, and I for each tone, to provide soft-output information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If MAP-based inner decoder is used, then bit-error-rate performance is optimized, but computational complexity becomes intractable as N and b increase
Solution Approach 1:
The patent segments the exhaustive search space of the MAP decoder into a tree structure with N levels, where each level corresponds to one of the N QAM symbols. By using a reduced-complexity M-algorithm, only M candidate paths are tracked at each level instead of exploring all possible paths, thereby segmenting the computational burden into manageable portions while maintaining near-optimal performance.
Solution Approach 2:
The patent applies partial action by using a fraction M of the total possible candidates at each tree level rather than examining all candidates. The parameter M is chosen to be much smaller than the total number of QAM constellation points, providing a practical compromise between computational feasibility and performance, achieving near-optimal bit-error-rate with significantly reduced complexity.
2Reliability
If MaxLogMAP-based inner decoder is used, then complexity is reduced and near-optimal performance is achieved at high SNR, but computational complexity still becomes intractable as N and b increase
Solution Approach 1:
The patent segments the computational task into a tree search with N levels, where each level processes one QAM symbol. The MaxLogMAP algorithm is applied at each level with only M candidate paths tracked, dividing the intractable full-search problem into N manageable sub-problems that can be solved sequentially with reduced complexity at each stage.
Solution Approach 2:
The patent uses partial action by tracking only M candidate paths at each tree level instead of all possible paths. This partial exploration provides a practical compromise that makes the computation tractable for large N and b while maintaining near-optimal performance, especially at high SNR conditions where the MaxLogMAP approximation is most effective.
3Device complexity
If QRD/M-Algorithm based inner decoder is used, then complexity is drastically reduced, but bit-error-rate performance deteriorates due to hard-output estimates
Solution Approach 1:
The patent introduces an intermediary soft-output mechanism that bridges the gap between the M-algorithm's hard decisions and the requirements for soft information in iterative decoding. By computing log-likelihood ratios based on the metrics accumulated during the M-algorithm tree search, the system generates soft outputs that enable effective iterative decoding, thereby improving bit-error-rate performance while maintaining the computational efficiency of the M-algorithm.
Solution Approach 2:
The patent implements feedback by using the soft-output information from the M-algorithm-based inner decoder to update the outer decoder, which in turn provides refined a priori information back to the inner decoder in subsequent iterations. This iterative feedback loop progressively improves bit-error-rate performance while maintaining the reduced complexity structure of the M-algorithm.
4Reliability
If iterative decoding is used with soft-output decoders, then performance is improved, but computational complexity increases
Solution Approach 1:
The patent segments the iterative decoding process into a fixed number of iterations between the inner M-algorithm decoder and outer code decoder. By limiting the number of iterations and using an efficient M-algorithm inner decoder with only M candidate paths, the overall computational complexity is controlled while still achieving the performance benefits of iterative decoding through multiple information exchanges between decoders.
Data Source
AI summary
A method and apparatus for tree position adaptive SOMA receiver structures are disclosed herein. In one embodiment, a device for use in a wireless communication system comprises a receiver to receive information-bearing signals from the transmitter wirelessly transmitted using OFDM and bit interleaved coded modulation, where the receiver includes an inner decoder structure having a soft output M-algorithm (SOMA) based multiple-in multiple-out (MIMO) joint demapper that uses a SOMA-based MIMO detection process to perform joint inner demapping over each tone, and wherein the SOMA-based MIMO joint demapper is operable to search a detection tree for each tone using a tree-search symbol order that is adapted for each tone based on channel state information and extrinsic information from the outer decoder, where only a number of best alternatives from every level of the tree are expanded.