Receiver Equalization Using Stochastic Hill Climbing and Genetic Mutation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing channel equalization techniques, such as least mean squares (LMS) and brute force search, face challenges in optimizing multiple parameters efficiently, leading to suboptimal solutions and prolonged equalization times due to coupling of parameters and excessive computational requirements.
Innovation Solution
The method employs stochastic gradient hill climbing and genetic mutation operations to optimize parameters for equalization, determining values that maximize the figure of merit of an eye diagram, allowing for efficient adaptation of both receivers and transmitters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If brute force search is used to find optimal parameter combinations, then manufacturing precision of equalization is improved, but loss of time becomes prohibitive
Solution Approach 1:
The patent segments the equalization process into two distinct phases: a training phase where exhaustive search is performed offline to determine optimal parameter combinations, and a real-time phase where pre-computed lookup tables are used for rapid equalization. This segmentation allows the computationally intensive search to be performed when time is not constrained, while real-time operation benefits from the segmented, pre-processed results.
Solution Approach 2:
The patent performs preliminary action by pre-computing optimal parameter combinations during an offline training phase and storing them in lookup tables. This preliminary computation eliminates the need for time-consuming searches during real-time equalization, as the system can directly retrieve pre-determined optimal parameters from the lookup tables based on measured channel characteristics.
2Device complexity
If multiple parameters are optimized using a single shared cost function, then device complexity is reduced, but manufacturing precision deteriorates due to parameter coupling
Solution Approach 1:
The patent segments the parameter optimization process by treating different equalization parameters (such as CTLE parameters and equalizer tap coefficients) as distinct optimization targets with separate cost functions. This segmentation prevents parameter coupling that would occur with a single shared cost function, allowing each parameter to be optimized independently to its true optimum without being constrained by suboptimal values of other parameters.
Solution Approach 2:
The patent changes the optimization approach by introducing multiple independent cost functions, each tailored to specific parameters or parameter groups. This allows the system to evaluate and optimize different parameters based on their specific impact on signal quality, rather than relying on a single aggregate cost function that may not accurately reflect the optimal settings for individual parameters.
3Manufacturing precision
If exhaustive search of all parameter combinations is performed, then manufacturing precision is improved, but productivity decreases due to extended equalization time
Solution Approach 1:
The patent performs preliminary action by conducting exhaustive searches during an offline training phase to build comprehensive lookup tables containing optimal parameter combinations for various channel conditions. This preliminary computation stores the results of what would otherwise be time-consuming searches, enabling rapid retrieval during real-time operation without sacrificing the precision that exhaustive search provides.
Solution Approach 2:
The patent segments the equalization system into an offline training component that performs exhaustive searches to generate lookup tables, and an online operational component that uses these pre-computed tables for rapid parameter selection. This segmentation allows the system to achieve both high precision (through exhaustive search in the training phase) and high productivity (through fast lookup in the operational phase).
Data Source
AI summary
A receiver receives communications over a communication channel, which may distort an incoming communication signal. In order to counter this distortion, the frequency response of the receiver is manipulated by adjusting several frequency response parameters. Each frequency response parameter controls at least a portion of the frequency response of the receiver. The optimal values for the frequency response parameters are determined by modifying an initial set of values for the frequency response parameters through one or more of stochastic hill climbing operations until a performance metric associated with the receiver reaches a local maximum. The modified values are displaced through one or more mutation operations. The stochastic hill climbing operations may subsequently be performed on the mutated values to generate the final values for the frequency response parameters.