MIMO Signal Decoding With Separate I/Q Detection for Lower Complexity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current MIMO techniques for wireless communication systems, particularly in W-LANs, face challenges with high computational complexity, especially when combined with high-order modulation schemes like 64-QAM, which limits their practical application due to increased latency and memory constraints.
Innovation Solution
The method simplifies the turbo-MIMO-MMSE reference scheme by exploiting Gray coding and matrix algebra properties to reduce the number of computations in key blocks like the soft-interference estimator, MMSE detector, and QAM soft de-mapper, without introducing approximations, and approximates the log-likelihood ratio calculation to lower overall system complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the MAP algorithm is used for optimal detection in MIMO systems, then detection performance is maximized, but computational complexity grows exponentially with modulation order and number of antennas
Solution Approach 1:
The patent segments the detection process into iterative steps using turbo-MIMO-MMSE architecture, where the detector and decoder operate in separate iterations. This divides the computationally intensive MAP algorithm into manageable segments that can be executed sequentially, reducing peak computational complexity while maintaining overall detection performance.
Solution Approach 2:
The patent changes the detection parameter from optimal MAP to sub-optimal MMSE, accepting a controlled performance trade-off to achieve polynomial rather than exponential computational complexity. This parameter change enables practical implementation in high-order modulation systems like 64-QAM with multiple antennas.
2Reliability
If MMSE detectors are used in iterative decoding schemes, then computational complexity increases proportionally with the number of iterations, impacting latency constraints
Solution Approach 1:
The patent performs preliminary computations of the correlation matrix and its inverse outside the iterative loop, caching these results for reuse across iterations. This preliminary action eliminates redundant calculations in each iteration, reducing the computational burden and latency while maintaining detection accuracy.
Solution Approach 2:
The patent implements a limited number of iterations (e.g., 2-3 iterations) rather than pursuing convergence to optimal performance. This partial action achieves sufficient system reliability for practical applications while significantly reducing the latency and computational overhead associated with extensive iterative processing.
3Productivity
If high-order modulation schemes like 64-QAM are used, then data rate increases, but computational complexity and memory requirements increase significantly
Solution Approach 1:
The patent segments the high-order modulation detection into iterative turbo-MIMO-MMSE steps, where each iteration processes a portion of the computational task. This segmentation makes high-order modulation feasible by distributing the computational load across multiple manageable iterations rather than requiring excessive resources in a single step.
Solution Approach 2:
The patent uses a limited number of iterations with simplified MMSE detection rather than exhaustive optimal detection. This partial action achieves acceptable performance for high-order modulation schemes like 64-QAM while keeping computational complexity and memory requirements at practical levels for implementation.
4Productivity
If the number of transmitting antennas is increased, then system performance and data rate improve, but computational complexity grows exponentially
Solution Approach 1:
The patent segments the MIMO detection problem into iterative turbo-MIMO-MMSE steps, where the increased complexity from multiple antennas is distributed across iterations. This allows the system to handle multiple transmitting antennas by processing the computational burden incrementally rather than all at once.
Solution Approach 2:
The patent changes from optimal MAP detection to sub-optimal MMSE detection, accepting a controlled performance trade-off to achieve polynomial computational complexity that scales manageable with the number of antennas. This enables practical implementation of multi-antenna systems with higher data rates.
Data Source
AI summary
A method and system for decoding signals includes a transmitter configured for transmitting signals encoded with a mapping, with different and separable configurations in a real part and an imaginary part of the signal. The signals may be encoded according to a Gray or QAM mapping, and may be transmitted on a selective MIMO channel and/or multiplexed with an OFDM technique. The corresponding receiver is configured for decoding the real part and the imaginary part of the signals separately, and may include a filter for subjecting the encoded signals to a Wiener filtering and a MMSE detector for minimizing the mean-square error between the encoded signals and the result of the Wiener filtering. The receiver may also include a soft decoder for performing a soft estimation of the signals and cancelling, using the results of the soft estimation, an interference produced on the signals.