Optimum frame synchronization over binary symmetric channel for data with known, unequal distribution

Based on hypothesis testing and using the likelihood ratio test (LRT), optimum frame synchronization metrics have been recently proposed for data transmission over an additive white Gaussian noise (AWGN) channel for frames which are of unknown, variable lengths. The derived metrics addressed the cases where data symbols have an equiprobable distribution, where the distribution of the data symbols is unknown, and where data symbols have a known, unequal distribution. In this paper, following a similar approach, a frame synchronization metric is proposed for transmission over a binary symmetric channel (BSC), where frames are of unknown, variable lengths and with data symbols which have a known, unequal distribution. The performance of this metric is compared via simulation with the commonly used hard correlation (HC) metric. The obtained receiver operating characteristic (ROC) curves show that an evident performance improvement can be obtained with the use of the proposed metric.