Noise Estimation in Modulated Data Carrier Signals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for estimating noise in wireless communication systems, such as signal-to-noise ratio (SNR) and Channel to Interference & Noise Ratio (CINR), suffer from biases and complexity, especially in interference-limited systems and fading channels, and often assume a Gaussian noise distribution, making them less effective.
Innovation Solution
A method that generates an instantaneous estimate of noise statistical properties from modulated data symbols using a compensation function, which is applied on a sample-by-sample basis and averaged to provide an unbiased estimation, suitable for various modulation schemes and noise distributions, including radial noise, without requiring decoding or encoding processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If hard slicing method is used to estimate CINR, then the estimation can be calculated from received and decoded points, but a bias of 2-7 dB is introduced due to decoding errors
Solution Approach 1:
The patent segments the noise estimation process into two distinct phases: a training phase where noise statistics are collected without decoding, and a working phase where the pre-estimated noise statistics are used. This segmentation eliminates the bias introduced by decoding errors in the estimation process itself, while still allowing decoding to occur for normal operation.
Solution Approach 2:
The patent performs preliminary noise estimation during a training phase before actual data decoding begins. By pre-characterizing the noise statistics (mean, variance, distribution type) without the interference of decoding operations, the method establishes accurate baseline parameters that are then used during the working phase, avoiding the 2-7 dB bias that plagues conventional hard slicing methods.
2Measurement precision
If decode-encode method is used for accurate CINR estimation, then the estimation accuracy is improved, but the implementation complexity increases and storage requirements increase
Solution Approach 1:
The patent extracts the noise estimation function from the main decode-encode process. Instead of requiring the full decode-encode cycle to estimate noise, the method extracts and utilizes only the received signal samples during a training phase to characterize noise statistics, then applies these pre-extracted parameters during working phase, eliminating the need for complex parallel encoding operations.
Solution Approach 2:
The patent creates a statistical model (copy) of the noise characteristics during the training phase, which then serves as a reference for all subsequent working phase operations. This statistical copy contains the mean, variance, and distribution parameters that are reused repeatedly, avoiding the need to perform computationally intensive decode-encode operations for each estimation.
3Ease of manufacture
If conventional noise estimation methods are used, then the process is simple, but the methods assume Gaussian noise distribution and are less suitable for interference-limited systems
Solution Approach 1:
The patent changes the parameters used for noise characterization from simple Gaussian assumptions to comprehensive statistical parameters including mean, variance, and distribution type identification. By estimating and adapting to the actual noise distribution parameters present in the channel, the method remains simple to implement while becoming versatile enough to handle non-Gaussian interference-limited conditions, fading channels, and mixed noise-interference environments.
Data Source
AI summary
Methods, systems and apparatus for estimating statistical properties of noise in modulated data carrier signals represented by modulated data symbols are disclosed. The method comprises generating an instantaneous estimate of the statistical property of noise from the received modulated data symbols on a sample by sample basis and applying a compensation function to the instantaneous estimate. The method further comprises averaging an output of the compensation function to determine the estimated statistical property of noise.