Multi-Branch RF Burst Equalization for CDMA Interference Cancellation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing wireless communication systems face challenges in effectively canceling co-channel and adjacent channel interference, particularly in CDMA downlink, where adaptive LMS algorithms produce biased signals and fail to sufficiently remove inter-symbol and inter-chip interference, leading to suboptimal signal processing.
Innovation Solution
The implementation of a multi-branch equalizer processing module that uses direct matrix inversion with a recursive algorithm like the Levinson algorithm for expeditious equalization training, followed by a second equalizer branch for improved interference cancellation, allowing for efficient processing of RF bursts and enhanced signal recovery in CDMA downlink and other wireless communication standards.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If adaptive LMS algorithms are used for interference cancellation, then the system is easier to implement, but the signal processing accuracy deteriorates due to biased signals and inability to adequately mitigate ISI and ICI
Solution Approach 1:
The patent divides the interference cancellation process into multiple stages: first using LMS algorithms for initial interference suppression, then applying direct matrix inversion to the residual signal for precise ISI and ICI mitigation. This segmentation allows each method to operate in its optimal performance range, combining implementation ease with high accuracy.
Solution Approach 2:
The patent introduces an intermediate processing stage where the output of the LMS algorithm serves as input to the direct matrix inversion algorithm. This intermediary approach allows the simpler LMS method to handle the bulk of interference cancellation while the more accurate matrix inversion method refines the result, achieving both ease of implementation and high precision.
2Measurement precision
If direct matrix inversion is used to train equalizers, then the training accuracy is improved, but the computational complexity increases
Solution Approach 1:
The patent applies direct matrix inversion selectively rather than throughout the entire processing chain. It uses matrix inversion specifically for training the equalizer and for mitigating residual ISI and ICI after LMS processing, while relying on the computationally simpler LMS algorithm for ongoing interference cancellation. This partial application maintains high accuracy where needed while controlling overall computational complexity.
Solution Approach 2:
The patent performs preliminary interference cancellation using LMS algorithms before applying direct matrix inversion. By removing the majority of interference in advance, the subsequent matrix inversion operates on a cleaner signal with reduced complexity requirements, while still achieving high training accuracy when needed.
3Device complexity
If LMS algorithms are used for interference cancellation, then the system complexity is reduced, but the ability to cancel co-channel and adjacent channel interference deteriorates
Solution Approach 1:
The patent segments the interference cancellation task into two parts: LMS algorithms handle general interference suppression with low complexity, while direct matrix inversion specifically targets residual co-channel and adjacent channel interference. This segmentation enables the system to maintain low overall complexity while achieving superior interference cancellation performance through the specialized matrix inversion stage.
Solution Approach 2:
The patent introduces direct matrix inversion as an intermediary processing stage between LMS cancellation and final signal recovery. This intermediary step specifically addresses the limitation of LMS algorithms in canceling co-channel and adjacent channel interference, enhancing the overall cancellation capability without significantly increasing system complexity since it operates only on residual interference.
Data Source
AI summary
The present invention provides a equalizer processing module operable to cancel interference associated with received radio frequency (RF) burst(s). This equalizer processing module includes a first equalizer processing branch and an optional second equalizer processing branch. The first equalizer processing branch is operable to be trained by applying a recursive DMI process such as a Levison algorithm, based upon known training sequences and equalize the received RF burst. This results in soft samples or decisions which in turn may be converted to data bits. The soft samples are processed with a de-interleaver and channel decoder, where the combination is operable to produce a decoded frame of data bits from the soft samples. This allows interfering signals to be cancelled and more accurate processing of the received RF bursts to occur.