Lattice Reduction for mmW Decoder Complexity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional demapping procedures in wireless communications, especially in millimeter wave (mmW) systems, are computationally complex and costly due to increased path loss and signal attenuation, leading to high power usage and large silicon die size, while less complex methods often result in worse decoding performance.
Innovation Solution
The use of lattice reduction (LR) techniques to generate a transformation matrix that can be reused across multiple resource elements, allowing for less computationally complex demapping and decoding methods like MMSE-based or SIC-based demapping, which provide similar performance to more complex methods by spreading the computational cost across multiple elements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional demapping procedures are used in mmW systems, then decoding performance is maintained, but computational complexity and power usage increase
Solution Approach 1:
The patent applies lattice reduction (LR) as a preliminary transformation to the channel estimate before demapping. This preprocessing step transforms the channel matrix to reduce interference between spatial layers, enabling simpler demapping algorithms to achieve performance comparable to complex conventional methods. The LR transformation is computed once and can be reused across multiple resource elements.
Solution Approach 2:
The patent changes the parameter representation of the channel by applying LR transformation to convert the original channel matrix into a transformed channel matrix with reduced interference characteristics. This parameter transformation enables the use of less complex demapping algorithms while maintaining decoding performance.
2Reliability
If conventional demapping procedures are used in mmW systems, then decoding performance is maintained, but power usage increases
Solution Approach 1:
The LR transformation is performed as a preliminary step that simplifies the subsequent demapping operation. By pre-transforming the channel estimate to reduce interference, the patent enables the use of lower-power demapping algorithms while maintaining decoding performance, thus reducing overall power consumption at the UE.
3Measurement precision
If lattice reduction is applied to each resource element individually, then decoding accuracy is maximized, but computational cost increases
Solution Approach 1:
The patent makes the LR transformation universal by applying it once to a reference resource element and then reusing the same transformation matrix across multiple resource elements. This approach maintains decoding accuracy for all REs while significantly reducing computational cost compared to performing LR individually for each RE.
Solution Approach 2:
The patent merges the LR operation across multiple resource elements by computing the transformation once and reusing it. This combines the computational effort into a single operation that serves multiple REs, reducing overall computational cost while maintaining accuracy through the robustness of beamformed transmissions to time-dispersion.
Data Source
AI summary
Methods, systems, and devices for wireless communications are described. Some wireless communications systems may utilize beamforming techniques to process wireless communications transmitted in millimeter wave (mmW) frequency ranges. In such cases, a user equipment (UE) may perform lattice reduction (LR)-based preprocessing for a received resource element (RE), which allows the UE to utilize demapping techniques (e.g., minimum mean square error (MMSE)-based demapping techniques or successive interference cancellation (SIC) demapping techniques) that are less computationally-complex than conventional demapping techniques (e.g., maximum likelihood (ML)-based demapping techniques) while providing a similar performance as conventional techniques. Further, due to mmW systems' robustness to time-dispersion, the UE may apply the same LR to multiple REs across multiple symbols in the time domain and across multiple sub-carriers in the frequency domain. The computational cost of performing the LR calculation may be spread across multiple REs and further increase the efficiency of utilizing low-complexity demapping techniques.