Noise Estimation Using Outer Constellation Points
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing noise estimation methods in communication systems suffer from inaccuracy due to mistaken evaluations of received symbols, leading to underestimation of noise power, especially when symbols are close to multiple constellation points, causing incorrect demodulation and distortion assessment.
Innovation Solution
A noise estimation apparatus and method that utilize only the outer constellation points for noise estimation, adjusting error calculation values based on the deviation zones to exclude mistaken evaluations, and scaling error values to compensate for noise power from inner constellation points, using a block diagram with evaluation, calculation, and mean calculation circuits to generate accurate noise estimation results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If all constellation points are used for noise estimation, then more data is available for calculation, but measurement precision deteriorates due to mistaken evaluations of received symbols
Solution Approach 1:
The patent extracts and excludes inner constellation points from noise estimation calculations, using only outer constellation points. This is achieved by determining the distance between received symbols and constellation points, identifying inner constellation points (those closer to the center of the constellation diagram), and excluding their error calculations from the noise power estimation. This resolves the contradiction by sacrificing quantity (fewer points used) to gain precision (avoiding mistaken evaluations from ambiguous demodulation near inner points).
2Ease of manufacture
If standard Euclidean distance calculation is used for all symbols, then calculation is simple, but noise power is underestimated due to mistaken symbol evaluation
Solution Approach 1:
The patent applies different quality treatment to different constellation points based on their positions. Outer constellation points are processed with standard Euclidean distance calculation, while inner constellation points are excluded from noise estimation. This local differentiation resolves the contradiction by maintaining calculation simplicity for valid points while improving overall accuracy through selective exclusion of problematic inner points that cause underestimation.
Solution Approach 2:
The patent segments the constellation points into two categories: outer constellation points and inner constellation points. This segmentation is based on calculating distances from the center of the constellation diagram and comparing against thresholds. By dividing the constellation points into segments with different roles in noise estimation, the patent resolves the contradiction between simple calculation and accurate measurement.
3Productivity
If received symbols near multiple constellation points are used, then more signal data is utilized, but reliability decreases due to ambiguous demodulation
Solution Approach 1:
The patent extracts and removes inner constellation points from the noise estimation process. These inner points are identified as those whose received symbols are likely to be ambiguously demodulated due to their proximity to multiple constellation points. By taking out these unreliable data points, the patent resolves the contradiction between utilizing more signal data and maintaining estimation reliability.
Data Source
AI summary
A noise estimation apparatus and method thereof includes an evaluation circuit, a first calculation circuit, a second calculation circuit, and a mean calculation circuit. The evaluation circuit is utilized for determining which constellation point on a constellation diagram each received symbol in a communication signal corresponds to for purposes of generating a relevant evaluated symbol and outputting an evaluated signal. The first calculation circuit is for generating an error output signal according to the communication signal and the evaluated signal, wherein the error output signal includes a plurality of error calculation values. The second calculation circuit is utilized for adjusting at least a portion of the plurality of error calculation values of the error output signal in order to output an adjusted error output signal. The mean calculation circuit is utilized for averaging the adjusted error output signal in order to generate a noise estimation result.