A novel noise floor estimation technique for optimized thresholding in spectrum sensing

Accurate Noise Floor Calibration based on Modified Expectation Maximisation of Gaussian Mixture

“…Notice that, in the most general case, the parameters of the Gaussian distribution (i.e., its mean and variance) are in general mutually independent (i.e., their values are unrelated). The general EM method assumes this general case where the mean and variance of the distribution are independent and therefore estimates both parameters independently from the sample set based on (13) and 14, respectively. However, in the Gaussian mixture considered in this work, the inspection of (3)-(6) reveals that mean and variance are not independent but related in closed form as:…”

“…Conversely, given a finite sample set, the sample estimates of lower order moments are in general more accurate. Therefore, a more accurate fit of (8)- (9) to P can be expected if the means {µ m } M m=0 (first raw moments) are first estimated from the sample set using (13) and the variances {σ 2 m } M m=0 (second central moments) are then estimated using (17). The proposed modified version of the EM method is thus obtained by replacing (14) with (17).…”

“…Most existing methods [10]- [13] rely on a threshold-based classification of the observed energy/power samples into two sets, one that is assumed to contain noise-only samples (i.e., samples of the observed energy/power when the PU is assumed to be absent) and another one that is assumed to contain signalplus-noise samples (i.e., samples of the observed energy/power when the PU is assumed to be present). The samples in the first set (i.e., the assumed noise-only samples) are then used to estimate the noise floor.…”

A Blind Signal Samples Detection Algorithm for Accurate Primary User Traffic Estimation

“…The two special operations of ROF are erosion-equivalent to lowest rank as it returns the minimum of the input set, and dilation-equivalent to highest rank as it returns the maximum. Erosion and dilation, besides being useful in image processing, can also be effectively used in impulse noise reduction and noise power estimation, as demonstrated in [9][10][11]. These studies iteratively increase the size of the filters used on the power spectrum samples.…”

“…In this paper, we present a practical algorithm for energy samples recognition-marking of signal and noise samples in a received time frame-of a dynamic PU signal. The algorithm uses rank order filtering, earlier studied for signal spectrum analysis only [9][10][11], for temporal signal analysis by redesigning the signal processing and samples marking processes. We evaluate the algorithm in terms of signal samples detection and complete samples recognition with respect to SNR and different PU activity factors, and also examine the execution time of the detection process.…”