Non-parametric Sign-based Spectrum Sensing for Heavy-tailed Noise
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional cyclostationarity-based spectrum sensing algorithms are not robust in the presence of heavy-tailed noise or interference, requiring a larger number of observations to maintain performance, especially when noise distributions deviate from Gaussian.
Innovation Solution
A non-parametric detector using a multivariate sign function for estimating cyclic correlation, which preserves cyclostationarity properties and achieves robust performance regardless of noise distribution, reducing the need for additional observations and nuisance parameter estimation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional cyclostationarity-based detectors are used, then detection capability is improved, but robustness against heavy-tailed noise deteriorates
Solution Approach 1:
The patent transforms the cyclic correlation estimator from a parametric form (assuming Gaussian noise) to a non-parametric form using sign functions. This parameter change allows the detector to maintain detection capability while becoming robust against heavy-tailed noise distributions, as the sign function bounds the influence of outliers.
Solution Approach 2:
The patent replaces the conventional correlation computation mechanism with a sign-based mechanism. Instead of computing actual correlation values that are sensitive to noise magnitude, the system uses signs of correlation values, which preserves detection capability while eliminating sensitivity to heavy-tailed noise characteristics.
2Reliability
If M-estimation techniques are used to improve robustness, then reliability is improved, but device complexity increases
Solution Approach 1:
The patent replaces complex iterative M-estimation procedures with a simple non-parametric sign-based estimator. This simpler 'disposable' estimator achieves similar robustness without requiring iterative computations or estimation of nuisance parameters, significantly reducing device complexity while maintaining reliability.
3Reliability
If more observations are collected to maintain performance in heavy-tailed noise, then reliability is improved, but loss of time increases
Solution Approach 1:
By changing the estimator parameterization to use sign functions, the patent achieves bounded influence of outliers, allowing reliable detection with fewer observations even in heavy-tailed noise environments, thereby reducing the time loss associated with collecting additional samples.
Data Source
AI summary
A user equipment senses a plurality of wireless signals it receives; estimates cyclic correlation of the received signals using a multi-variate sign function; determines from the estimated cyclic correlation which ones of the signals are detected based on cyclostationarity present at known cyclic frequencies; and a frequency resource is selected for opportunistic communications based on the frequencies over which were received the signals that were determined to be detected. In specific embodiments, the multi-variate sign function is bivariate; the detected signals are primary user signals and the selected frequency resource avoids frequencies on which they were received; and/or the detected signals are secondary user signals and the selected frequency resource can either avoid those frequencies on which they were received or use those frequencies for the UE's own opportunistic communications.