Iterative MIMO Channel Equalizer Using Segmented QR Decomposition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional iterative receivers for MIMO uplink systems are costly due to their complexity and resource-intensive implementation, particularly in terms of semiconductor chip size and power consumption, which limits their widespread adoption.
Innovation Solution
A method and apparatus for channel equalization in an iterative receiver that reduces complexity by performing a full QR decomposition only in the initial iteration and subsequent updates for each transmission antenna, using a single filter matrix and log-likelihood ratio information to refine user matrices, thereby minimizing computational resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a full QR decomposition is performed for each transmission antenna in every iteration, then channel equalization performance is improved, but device complexity and power consumption increase excessively
Solution Approach 1:
The patent segments the QR decomposition operation by separating the upper triangular matrix R (which remains constant across iterations) from the orthogonal matrix Q (which is updated). Only the necessary portions are recomputed in each iteration, dividing the computational task into static and dynamic components that can be handled differently.
Solution Approach 2:
The patent performs the complete QR decomposition once in advance to obtain the initial R matrix, which is then reused in subsequent iterations. This preliminary computation eliminates the need to repeatedly perform full decompositions, significantly reducing iterative computational complexity while maintaining equalization performance.
2Reliability
If a full QR decomposition is performed for each transmission antenna in every iteration, then channel equalization performance is improved, but power consumption increases excessively
Solution Approach 1:
The patent segments the computational workload by identifying that only certain matrix operations need to be performed in each iteration. By separating computations into one-time setup operations and iterative update operations, the patent reduces the energy-intensive operations to only what is necessary for performance improvement.
Solution Approach 2:
The patent performs energy-intensive QR decomposition operations once in advance before the iterative process begins. This preliminary computation shifts the high power consumption to an initial setup phase, allowing the iterative equalization to proceed with much lower power requirements, thus improving overall energy efficiency.
3Reliability
If multiple receiver processing chains are implemented, then iterative reception capability is achieved, but semiconductor chip size increases
Solution Approach 1:
The patent merges common computational resources across multiple receiver processing chains by identifying and sharing the R matrix computation and other invariant operations. This allows multiple chains to cooperate and reuse computational infrastructure, reducing the total chip area required compared to fully independent chains.
Solution Approach 2:
The patent creates universal computational blocks that can serve multiple receiver processing chains simultaneously. The QR decomposition engine and matrix operation units are designed to be multi-functional, handling computations for different chains in sequence or parallel with resource sharing, thereby reducing overall hardware footprint.
Data Source
AI summary
A method relates generally to channel equalization. In this method, a filter matrix is determined for transmission antennas by a channel equalizer of a first receiver processing chain. A first QR decomposition is performed on a first extended matrix for a first iteration. LLRs are fed from a second receiver processing chain to the first receiver processing chain for a second iteration. Symbol information is obtained from the LLRs. Interference is canceled using the symbol information to provide residual information. The channel equalizer is updated with the symbol information. The residual information is provided to the channel equalizer. User matrices corresponding to the transmission antennas are determined by the channel equalizer. This determination includes performing a second QR decomposition on a second extended matrix to obtain updated values for the user matrices, and performing updates using the symbol information and the updated values to provide the user matrices.