Using spectral correlation for non-cooperative RSS-based positioning

Abstract

Typically, methods for received signal strength (RSS) based source localization assume that the signal can be separated from the noise and interference. However, in non-cooperative applications where the details of the modulation format are not known a priori, this is not possible. In a high SNR environment, the RSS of the signal can be estimated from the power spectral density (PSD) via energy detection, but the PSD becomes dominated by noise at low SNR, leading to a loss of positioning information. This paper proposes the use of the spectral correlation function (SCF) for RSS-based localization, as the SCF has been shown to outperform simple energy detection for detecting the presence of a signal.