Hybrid blind symbol rate estimation for linearly modulated signals

Abstract

In this paper, a hybrid two-stage algorithm for blind symbol rate estimation is proposed, which applies for raised cosine pulse shaped, linearly modulated signals. In the first stage, the inverse (fast) Fourier transform (IFFT) of the averaged power spectrum is used to extract a coarse estimate based on the evaluation of the related baseband signal. In the second stage, a Maximum-likelihood (ML) based algorithm is used for the fine estimation due to its high accuracy. A prior estimation of the carrier frequency is proposed to allow for working in blind environment at low signal-to-noise ratio (SNR) levels. The performance of the proposed algorithm verified by simulation shows a high efficiency with less complexity comparing to the methods published in the open literature.