Signal Detection via Channel Matrix Factorization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current signal detection methods in communication systems, such as MIMO and OFDM-CDMA, face challenges with high complexity and poor performance, particularly with the LLL method, which lacks deterministic complexity and lacks a performance index for orthogonalization, making it impractical for hardware implementation and ineffective in actual systems.
Innovation Solution
A signal detecting method that factorizes the channel matrix into a more manageable form using a factorization matrix D, allowing for reduced error estimation by transforming the channel matrix H into {tilde over (H)}D, and selecting a detection matrix G to minimize the mean square error, thereby improving detection efficiency and error rate performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If maximum likelihood detection is used, then detection performance (error rate) is improved, but computational complexity increases exponentially
Solution Approach 1:
The patent segments the detection process by factorizing the channel matrix H into H = ÷D, where à is a simplified channel matrix and D is a factorization matrix. This segmentation allows the detection to be performed in two stages: first transforming the received signal through D, then detecting through the simplified matrix Ã, thereby reducing the exponential complexity while maintaining detection performance.
Solution Approach 2:
The factorization matrix D serves as an intermediary that transforms the original detection problem into a simpler form. By introducing this intermediate transformation, the patent converts the complex direct detection through H into a two-step process that reduces computational burden while preserving the essential detection accuracy.
2Device complexity
If LLL method is used, then computational complexity is reduced, but detection performance becomes too poor to be accepted
Solution Approach 1:
The patent changes the approach by introducing a factorization matrix D that transforms the channel matrix into a form suitable for efficient detection. This parameter change enables the system to achieve both low complexity and acceptable performance by optimizing the factorization rather than using fixed methods like LLL.
Solution Approach 2:
The patent employs feedback mechanisms to determine the optimal factorization matrix D. By iteratively adjusting D based on performance criteria and using the received signal characteristics to refine the factorization, the system achieves both low complexity and high detection performance.
3Device complexity
If LLL method is used, then detection complexity is reduced, but hardware implementation becomes difficult due to nondeterministic complexity
Solution Approach 1:
The patent introduces a dynamic factorization process where the matrix D can be adjusted based on actual signal conditions. This dynamic approach allows the system to adapt to different channel conditions while maintaining deterministic and predictable complexity, making hardware implementation feasible.
Solution Approach 2:
The patent performs preliminary factorization of the channel matrix into H = ÷D before the actual detection process. This preliminary action simplifies the subsequent detection steps and provides a deterministic complexity profile that is easy to implement in hardware, avoiding the nondeterministic nature of methods like LLL.
4Device complexity
If LLL method is used, then detection complexity is reduced, but no performance index is provided for measuring orthogonalization
Solution Approach 1:
The patent replaces the mechanical iterative process of LLL with a matrix factorization approach that provides explicit performance metrics. By using the factorization matrix D and simplified channel matrix Ã, the system achieves reduced complexity while introducing clear performance indices for measuring detection accuracy and orthogonalization.
Data Source
AI summary
A signal detecting method and a receiver using the same are provided. The method includes the following steps. A receiving signal vector y is received through a number of channels, wherein the receiving signal vector y corresponds to a transmitting signal vector x transmitted by at least one of the channels. A channel matrix H is determined, wherein the channel matrix H represents at least one of the channels. A factorization matrix D is chosen, wherein D is invertible to make the channel matrix H expressed as H={tilde over (H)}D, and {tilde over (H)} is a corresponding channel matrix. The factorization matrix D is determined to make an expected value of the signal estimate error become smaller. The receiving signal vector y is detected to estimate the transmitting signal vector x according to the corresponding channel matrix {tilde over (H)} and the factorization matrix D.