Noise Variance Estimation With Narrow-Band Interference Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Digital communication systems, such as MoCA networks, face challenges in achieving coding gains in the presence of both additive white Gaussian noise and interference, which affect bit error rates and overall system performance.
Innovation Solution
The system employs a combination of long-term data-aided and non-data-aided algorithms to estimate noise variance and interference variance, determining an interference hypothesis based on these estimates to improve log-likelihood ratio computations and bit error rate performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional noise variance estimation methods are used, then the estimation process is simple, but the bit error rate performance deteriorates in the presence of narrow-band interference
Solution Approach 1:
The patent segments the noise variance estimation process into two distinct phases: a training phase using known training symbols to establish baseline noise characteristics, and a data phase using decision-directed error calculations to track noise variations. This segmentation allows the system to maintain simple per-subcarrier estimation while achieving robust performance through phased processing
Solution Approach 2:
The patent performs preliminary noise variance estimation during the training phase before actual data transmission begins. By pre-characterizing the noise environment using known training symbols and calculating initial per-subcarrier noise variances, the system establishes a foundation for more accurate subsequent decoding without adding complexity to the main data processing path
2Reliability
If per-subcarrier noise variance estimation is performed, then coding gains are improved in noisy environments, but the computational complexity increases
Solution Approach 1:
The patent applies local quality by estimating noise variance independently for each subcarrier rather than using a single global noise variance value. This allows the system to adapt to frequency-selective fading and localized interference conditions, providing optimal decoding performance for each subcarrier according to its specific noise characteristics
Solution Approach 2:
The patent implements a decision-directed approach where the decoder uses its own decoded symbols to calculate estimation errors and update noise variance estimates. This self-service mechanism allows the system to continuously refine noise characteristics without external assistance or additional training sequences, reducing overall system complexity
3Reliability
If interference detection algorithms are added to improve performance in interfering environments, then the bit error rate performance improves, but the device complexity and processing overhead increase
Solution Approach 1:
The patent merges interference detection and noise variance estimation into a unified process. By calculating the magnitude of received training symbols and comparing against thresholds during the same processing pass as noise estimation, the system detects narrow-band interference without requiring separate detection algorithms or additional processing overhead
Solution Approach 2:
The patent creates a universal noise variance estimation framework that simultaneously handles multiple functions: estimating thermal noise, detecting narrow-band interference, and providing per-subcarrier variance values for decoding. This multi-functional approach eliminates the need for separate specialized algorithms for each function
Data Source
AI summary
Method and apparatus are provided for estimating noise variance using a long-term data aided algorithm and an interference variance using a short-term data aided algorithm. Using these estimations, an interference hypothesis may be determined. Some embodiments compute the variance for a decision directed noise sample, convert the variance for the decision directed noise to a true noise variance per packet, and convert the variance for the decision directed noise to a true noise variance per symbol. The interference hypothesis may be based on the noise variance estimations per symbol and the noise variance estimation per packet. Some embodiments determine the presence of noise based on a comparison of the noise variance per packet, the noise variance per symbol, and each hypothesis and compute the long-term noise variance. Using the long-term noise variance in place of the variance for a decision directed noise sample for a subsequent determination of interference.