Robust and accurate frequency and timing synchronization using chirp signals

We propose a new robust and accurate synchronization procedure using a training sequence composed of chirp signals. We use a new integer frequency estimation algorithm and propose a new combination of a known fractional frequency offset estimation algorithm and timing synchronization algorithm. The training sequence is composed of one up and two down chirp symbols, also known as Newman phases. The integer frequency offset estimation algorithm uses the effect of timing and frequency offsets on the matched filter outputs of the chirp signals. Autocorrelation and reversed autocorrelation are used to acquire the timing instant and the fractional frequency offset. We present the complete timing and frequency synchronization procedure and study the output signals of the autocorrelation and reversed autocorrelation algorithms. Finally, we check the performance of the synchronization procedure via Monte Carlo simulation in several multipath channels. Our algorithms are accurate and more robust compared to previously published state-of-the art algorithms.