Maximum Likelihood Timing and Carrier Synchronization in Burst-Mode Satellite Transmissions

Michele Morelli,Antonio A D'Amico

doi:10.1155/2007/65058

Abstract

This paper investigates the joint maximum likelihood (ML) estimation of the carrier frequency offset, timing error, and carrier phase in burst-mode satellite transmissions over an AWGN channel. The synchronization process is assisted by a training sequence appended in front of each burst and composed of alternating binary symbols. The use of this particular pilot pattern results into an estimation algorithm of affordable complexity that operates in a decoupled fashion. In particular, the frequency offset is measured first and independently of the other parameters. Timing and phase estimates are subsequently computed through simple closed-form expressions. The performance of the proposed scheme is investigated by computer simulation and compared with Cramer-Rao bounds. It turns out that the estimation accuracy is very close to the theoretical limits up to relatively low signal-to-noise ratios. This makes the algorithm well suited for turbo-coded transmissions operating near the Shannon limit.

Highlights

Burst transmission of digital data and voice is widely adopted in satellite time-division multiple-access (TDMA) networks
In these applications the propagation medium can be reasonably modeled as an additive white Gaussian noise (AWGN) channel and knowledge of carrier frequency, symbol timing, and carrier phase is necessary for coherent demodulation of the received waveform
We report on simulation results illustrating the performance of ML estimator (MLE) over an AWGN channel

Summary

INTRODUCTION

Burst transmission of digital data and voice is widely adopted in satellite time-division multiple-access (TDMA) networks In these applications the propagation medium can be reasonably modeled as an additive white Gaussian noise (AWGN) channel and knowledge of carrier frequency, symbol timing, and carrier phase is necessary for coherent demodulation of the received waveform. The problem of timing and frequency offset recovery still remains Depending on their topology, synchronization circuits can be divided into two main categories: feedback and feedforward schemes [1, 2]. The common approach to solve the synchronization problem in burst-mode transmissions is to estimate the timing error first, and use the time-synchronized samples for frequency and phase recovery.

Statement of the problem

Signal model

Maximization of the likelihood function

CRB ANALYSIS

SIMULATION RESULTS

CONCLUSIONS