Radio signals detection thresholds estimation for energy detection technique.

Abstract

The estimation method for signal-to-noise ratio threshold values pertinent for low-level signals detection is developed and employed. The method is applicable for original energy detection technique based upon distribution-free statistics computed using spectrum samples provided by measuring equipment. Estimated thresholds are determined by measuring equipment inherent noise fluctuations and can be established in advance for certain hardware settings and sample lengths. For a typical spectrum analyzer model estimated sensitivity threshold varied from +0,6 dB to –11 dB for spectrum samples lengths ranged between 470 and 30 000 spectrums respectively. Experimental data confirmed estimated values and equivalence of sensitivity thresholds for white noise (generated by analog generator) and broadcasted LTE signals (generated by cellular base stations). The suggested energy detection technique is independent of signal's modulation, signal's probability distribution features, and intermittent or sporadic signal's total duration allotment profile within data acquisition period.