Another turbo-something : Carrier synchronization

We introduce in this paper a low-complexity joint carrier frequency and phase recovery algorithm for coherent detection/decoding of a turbo-coded 16-QAM signal. The estimator (which can be actually applied to any linear modulation scheme) is based on a pseudo-Maximum Likelihood (ML) approach, and unlike previously published works, makes iterative use of soft decisions provided by the SISO (Soft-In Soft-Out) decoders within the overall iterative turbo decoding scheme, yielding negligible degradation with respect to ideal carrier synchronization. Performance in terms of mean estimated value, mean-squared estimation error, and overall decoder Bit Error Rate (BER) as derived by simulation are also reported.

1. Introduction

The impressive performance of turbo codes has triggered in the last decade a lot of research addressing the application of this powerful coding technique in digital wireless communications [1]–[3]. When applied to linear modulations, in order to reach a performance which is close to Shannon capacity on the AWGN channel, one has to implicitly assume ideal coherent detection, or, in other words, the carrier reference has to be estimated before the data can be decoded. Therefore, the problem is how to achieve a fast and accurate carrier synchronization (due to the instability of the oscillators and the Doppler effect) from the received signal at those extremely low signal-to-noise ratios typical of such codes.