In this paper novel flipped parameter technique (FPT) for time delay estimation (TDE) in source localization problem is described. We propose passive source localization technique based on the development of an energy efficient algorithm that can reduce intersensor and interarray communication. We propose a flipped parameter (FP) which can be defined for any sensor in distributed sensor subarrays during the observation period. Unlike classical TDE methods that evaluate cross-correlation function, FPT requires evaluation based upon single sensor signal. The computed cross correlation between a signal and its analytic 'flipped' pair (flipped correlation) is a smooth function which peak (time delay) can be accurately detected. Flipped parameters are sufficient to determine all differential delays of the signals related to the same source. The flipped parameter technique can be used successfully in two-step methods of passive source localization with significantly less energy in comparison to the classic cross correlation. The use of FPT method is especially significant for the energy constrain distributed sensor subarrays. Using synthetic seismic signals, we illustrate the error of the source localization for classical and proposed method in the presence of noise. We demonstrate the performance improvement in noise environment of the proposed technique in comparison to the classic methods that use real signals. The proposed technique gives accurate results for both coherent and non-coherent signals.
Spherical-Harmonic-Domain Feedforward Active Noise Control Using Sparse Decomposition of Reference Signals from Distributed Sensor Arrays