Signal Parameter Estimation Using Spectral Correlation Triplets
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for estimating signal parameters in a frequency band with high levels of interference are ineffective, especially when the Signal to Interference plus Noise Ratio (SINR) is low, and fail to exploit the amplitude of spectral peaks, making it difficult to separate useful signals from interfering signals.
Innovation Solution
A method involving the use of the spectral correlation function (SCF) and cyclic autocorrelation function to extract peaks and identify signature triplets, which allows for the estimation of signal parameters without prior knowledge of interfering signals, by processing the signal through various orders of transformation and exploiting both the positions and amplitudes of peaks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If classic signal estimation methods are used, then the method is simple to implement, but the estimation accuracy is highly degraded when interfering signals are present at high levels
Solution Approach 1:
The patent segments the complex signal estimation problem into multiple processing stages: first applying cyclic autocorrelation to extract periodic characteristics, then applying spectral correlation to identify peak patterns, and finally using triplet identification to separate useful signals from interferers. This multi-stage segmentation allows accurate parameter estimation even when interfering signals are present at high levels comparable to the useful signal.
2Measurement precision
If spectral correlation function with multiple transformations is used, then signal parameter estimation accuracy is improved, but the computational complexity increases
Solution Approach 1:
The patent applies preliminary action by first computing the cyclic autocorrelation function to identify candidate peaks before proceeding to spectral correlation analysis. This preliminary step filters out non-promising frequency components, reducing the computational burden of subsequent spectral correlation calculations while maintaining accurate signal parameter estimation.
Solution Approach 2:
The patent applies local quality by focusing computational resources on specific regions of the spectrum where signal peaks are identified. Rather than processing the entire frequency spectrum uniformly, the method concentrates analysis on local peak regions and their neighborhoods, extracting triplet information only where relevant signals are present, thereby reducing overall computational complexity.
3Measurement precision
If prior knowledge of interfering signals is required, then parameter estimation can be performed cooperatively, but the method becomes inapplicable when no information is available about interferers
Solution Approach 1:
The patent implements self-service by enabling the signal estimation system to automatically identify and characterize both useful signals and interfering signals without requiring prior knowledge or external information about the interferers. The cyclic autocorrelation and spectral correlation functions automatically reveal the structure and parameters of all present signals, allowing the system to adaptively separate and estimate parameters of the useful signal even in non-cooperative environments where interferers are completely unknown.
Data Source
AI summary
A method for estimating parameters of a single signal or mixed-frequency signals in a given frequency band comprises: receiving the signal(s); calculating a spectral correlation function having variables f and α (cyclic frequency) for different values of L taken from a set of q strictly positive values, where, for each q, α can independently take on one or more discrete values and/or cover one or more value ranges; from each calculation, extracting a set of local maxima detected on the basis of the variables f and α or on the basis of variable f for each discrete value of α, each of said peaks being characterized by a triplet [LApi, LFpi, Lαpi], LFpi and Lαpi being the frequency and cyclic frequency of said triplet, respectively, and LApi being the amplitude; and identifying groups of triplets as being the signature left by a set of parameters making up the signal(s).