Multi-stream Demodulation via Non-ML Candidate Generation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In multi-antenna wireless communication systems, existing demodulation techniques are complex and resource-intensive, particularly in MIMO systems, where they struggle to efficiently handle varying channel conditions and high data rates, leading to increased processing complexity and time.
Innovation Solution
The implementation of a multi-stream demodulation method using non-maximum likelihood (non-ML) detection techniques to generate candidate vectors, followed by selection using maximum likelihood (ML) or hard/soft ML techniques, reduces complexity by considering only a subset of candidate vectors, thereby simplifying the demodulation process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If maximum likelihood (ML) detection techniques are used to handle varying channel conditions and high data rates in MIMO systems, then demodulation performance is improved, but processing complexity and time increase significantly
Solution Approach 1:
The patent segments the complex ML detection problem into two stages: first, generate a limited set of candidate vectors using simplified non-ML detection techniques; second, apply ML detection only to these candidates. This segmentation reduces the search space from all possible transmitted vectors to a manageable subset, thereby reducing processing complexity while maintaining demodulation performance.
Solution Approach 2:
The patent applies partial action by using non-ML detection techniques to generate candidate vectors that are likely to contain the transmitted vector, rather than exhaustively checking all possible vectors. This partial approach (checking only promising candidates) reduces computational complexity while still achieving reliable demodulation performance.
2Measurement precision
If exhaustive ML detection is performed on all possible candidate vectors, then demodulation accuracy is maximized, but processing time increases
Solution Approach 1:
The patent performs preliminary action by using non-ML detection techniques to pre-generate and filter candidate vectors before applying ML detection. This preliminary filtering step eliminates obviously incorrect candidates, so that ML detection is applied only to a reduced set of promising candidates, thereby reducing processing time while maintaining demodulation accuracy.
Solution Approach 2:
The patent applies local quality by using different detection approaches for different parts of the problem: simplified non-ML techniques for generating candidates (where high accuracy is less critical) and full ML detection for the final selection (where accuracy is critical). This localized application of different methods optimizes the balance between accuracy and processing time.
3Adaptability or versatility
If multiple detectors are used to generate candidate vectors, then coverage of possible transmitted vectors is improved, but device complexity increases
Solution Approach 1:
The patent applies universality by designing detectors that can operate in multiple modes or serve multiple functions. The detectors generate candidate vectors that cover various possible transmitted signals, and the same detector structure can be reused with different parameters or configurations to handle different channel conditions, reducing the need for entirely separate detector systems.
Solution Approach 2:
The patent implements nesting by having multiple detectors operate hierarchically, where each detector generates a subset of candidate vectors that are then combined and further processed. The output of one detector becomes input to another, creating a nested structure that achieves comprehensive coverage through coordinated operation of simpler components rather than requiring all detectors to operate independently at full complexity.
Data Source
AI summary
In a method of demodulating a signal transmitted via a multiple input multiple output (MIMO) communication channel, a data symbol vector is received at a communication device, the data vector comprising a plurality of data symbols that are received via a plurality of antennas. The received data symbol vector corresponds to a transmitted data symbol vector comprising a plurality of transmitted data symbols. A plurality of candidate vectors for the transmitted data symbol vector are generated at the communication device using a plurality of non-maximum likelihood detection techniques. One candidate vector is selected from the plurality of candidate vectors at the communication device using a maximum likelihood technique.