Primary Signal to Noise Ratio Estimation Based on AIC for UWB Systems

Abstract

Ultra-Wide-Band (UWB) devices need detect and avoid techniques in order to avoid or reduce interference to primary systems whose spectra overlap bands of the UWB systems. Some avoidance techniques require a knowledge of signal level received from the primary systems to control the transmitted power. Thus, detection schemes have to accurately estimate the primary signal level using the observed signal includes an additive noise and to provide it for the avoidance schemes. In this paper, we propose a new method to estimate the Primary Signal to Noise Ratio (PSNR) for the detection scheme. Our proposed method uses the fast Fourier transform output of a Multi-Band Orthogonal Frequency Division Multiplexing system. We generate models based on whether the primary signals are present, estimate the PSNR using a maximum likelihood criterion in each model and obtain the PSNR estimate by selecting the most preferable model using an Akaike information criterion. The propose method does not need any a priori information of the primary signal and the additive noise. By computer simulations, we evaluate an accuracy of the PSNR estimation of the proposed method.