MIMO Weight Vector Continuity via Common Autocorrelation Matrix
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In MIMO eigenmode systems, maintaining continuity of weight vectors in the frequency domain is challenging, leading to difficulties in estimating channels with delay spread beyond the guard interval, which affects signal-to-noise ratio and receiving characteristics.
Innovation Solution
A method involving a transmitting apparatus that acquires a channel matrix, derives a common autocorrelation matrix, and uses eigenvalue decomposition to generate a steering matrix, performing Gram-Schmidt orthonormalization to derive weight vectors that ensure continuity across subcarriers, enabling effective transmission of multicarrier signals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If MIMO eigenmode system uses eigenvalue operation per subcarrier to form eigenbeams, then channel capacity is increased, but continuity of weight vectors in frequency domain deteriorates
Solution Approach 1:
The patent segments the weight vector derivation process into two parts: a common autocorrelation matrix derived from multiple channel matrices (providing frequency domain continuity) and individual channel-specific steering matrices (providing eigenbeam formation capability). This segmentation allows each part to fulfill its specific function while maintaining overall system performance.
Solution Approach 2:
The patent merges the common autocorrelation matrix (which ensures frequency continuity) with individual channel matrices (which capture frequency-specific characteristics) to create a hybrid weight vector derivation approach. This combination preserves the benefits of both continuous weighting and eigenmode optimization.
2Object-affected harmful factors
If smoothing processing is performed in frequency domain, then noise effect is reduced, but orthogonal channel formation is prevented in MIMO eigenmode system
Solution Approach 1:
The patent shifts the smoothing operation from the frequency domain to the spatial domain by applying it to the common autocorrelation matrix before eigenvalue decomposition. This dimensional change allows noise reduction without interfering with the orthogonal channel formation process in the frequency domain.
3Productivity
If weight vectors lack continuity in frequency domain, then eigenbeam formation is achieved, but delay spread beyond guard interval increases
Solution Approach 1:
The patent performs preliminary derivation of the common autocorrelation matrix from multiple channel matrices before forming the final weight vectors. This preliminary action captures the statistical continuity across frequencies, which then constrains the final weight vectors to maintain continuity and limit delay spread.
Data Source
AI summary
A plurality of antennas transmit multicarrier signals composed of a plurality of streams. An IF unit acquires, per carrier, a channel matrix having elements the number of which is determined by the number of a plurality of transmitting antennas and the number of a plurality of receiving antennas provided in a receiving apparatus. A baseband processing unit derives a common autocorrelation matrix for the channel matrix acquired per carrier. The baseband processing unit derives a steering matrix by eigenvalue-decomposing the derived common autocorrelation matrix. The baseband processing unit derives, per carrier, weight vectors for a plurality of streams in a manner such that an orthonormalization is performed respectively on matrices obtained by operating the derived steering matrix on the channel matrix per carrier.