Enhanced SIC Receiver for MIMO Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current two-stream receivers for MIMO systems face challenges in achieving low-latency and low-complexity processing while maintaining spectral efficiency, with existing methods like brute force maximum likelihood, Deterministic Sequential Monte-Carlo, and successive interference cancellation suffering from high complexity, sub-optimal demodulation, and error propagation.
Innovation Solution
The enhanced successive interference cancellation (SIC) receiver method lists all possibilities for symbols, determines exact maximum log likelihood ratios, decodes codewords, and reconstructs streams with error correction, reducing complexity and error propagation by subtracting only correctly decoded codewords, and extends to multiple streams using modified QR decomposition and MMSE filtering.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If brute force maximum likelihood reception method is used, then optimal demodulation is achieved, but device complexity becomes highly complex
Solution Approach 1:
The patent segments the two-stream reception problem into two separate single-stream decoding stages. First, codeword-1 is decoded from the received signal, then its contribution is subtracted to decode codeword-2. This segmentation transforms the exponentially complex joint decoding into two manageable sequential decoding operations, reducing complexity while maintaining near-optimal performance through successive interference cancellation.
Solution Approach 2:
The patent applies preliminary action by computing exact max-log LLRs for codeword-1 before decoding, using all possible symbol pairs (s1, s2) to determine accurate soft outputs. This preliminary computation of reliable LLRs ensures that the first decoding stage operates with optimal information, improving the likelihood of successful decoding and reducing error propagation to the second stream.
2Device complexity
If Deterministic Sequential Monte-Carlo method is used, then complexity is reduced, but demodulation becomes sub-optimal due to missing candidate problem
Solution Approach 1:
The patent extracts and eliminates the source of the missing candidate problem by using exact max-log LLR computation that considers all possible symbol pairs. Instead of relying on a reduced hypothesis set that may miss candidates, the invention extracts the complete information from all (s1, s2) pairs through the metric evaluation, ensuring no candidate is missed while maintaining computational efficiency through the sequential decoding structure.
3Device complexity
If conventional SIC receiver is used, then complexity is reduced, but error propagation occurs due to Gaussian interferer assumption
Solution Approach 1:
The patent changes the fundamental parameter assumption in SIC by replacing the Gaussian interferer approximation with exact max-log LLR computation. Instead of assuming the interfering stream follows a Gaussian distribution, the invention computes precise likelihood ratios by evaluating metrics over all possible symbol pairs, fundamentally changing the statistical parameter model to achieve accurate soft outputs without error propagation.
Solution Approach 2:
The patent implements feedback by using the decoded codeword-1 to compute its contribution to the received signal and subtracting it from the original received signal before decoding codeword-2. This feedback loop ensures that the decoded information is actively used to cancel interference, improving the signal quality for the second decoding stage and reducing error propagation through iterative refinement.
Data Source
AI summary
A method includes the steps of: i) listing out all possibilities for first symbol of a two stream signal; ii) determining a second symbol of the two stream signal for each said first symbol listed out, iii) evaluating a metric for each said first symbol and second symbol pair, iv) determining the exact maximum log likelihood ratio for all bits associated with said first symbol using said metrics, v) decoding a codeword-1 using the maximum log likelihood ratios, vi) re-encoding said codeword-1, vii) modulating said re-encoded codeword-1, viii) subtracting said modulated re-encoded codeword-1 from said two stream signal, ix) determining metrics for all possibilities for second symbol in the signal obtained in viii, x) determining the maximum log likelihood ratios for all bits associated with second symbol, and xi) decoding said codeword-2 using the maximum log likelihood ratios for all bits associated with said second symbol.