Extrinsic Information Aided Frame Synchronizer for Turbo Code

The available soft outputs of the turbo decoders are tried to help the frame synchronizers in the form of mean of the extrinsic information's absolute value (MEIAV). It is found that the means and standard deviations of the MEIAV after the first iterative step when frame synchronization is perfectly achieved are distinct from those that are obtained when frame synchronization loses. And these statistics also give some hints to explain the relationship between the frame synchronization error rate (FSER) performance and some parameters such as interleaver size and constraint length of constituent codes. In order to further lower the complexity and improve FSER performance of the extrinsic information based frame synchronization algorithm, a new frame synchronizer for turbo coded systems based on the combination of maximum a posteriori (MAP) belief metrics from turbo decoder and maximum likelihood (ML) belief metrics from attached synchronization maker (ASM) correlation is proposed in this paper. Simulation results show that the new algorithm can achieve much lower FSER and consequently bring about some additional code gains in BER and FER performance. And the new synchronizer has less calculated load than the extrinsic based scheme.