Design of Efficient Convolutional and Serially Concatenated Convolutional Codes with A-Priori Information

Abstract

In this paper, we focus on the design of optimized binary convolutional codes (CCs) and serially concatenated convolutional codes (SCCCs) in the presence of a-priori information (API) at the receiver. First, we propose a design criterion for CCs based on the minimization of the bit error proabability (BEP). In this case, relevant performance gains, with respect to previously proposed CCs, are obtained. These gains persist even in the presence of estimation errors on the API. Then, we apply the same BEP-based design criterion to SCCCs and derive good encoders' structures. In the SCCC case with API at the receiver, simulation results show substantial gains with respect to previously proposed parallel concatenated convolutional coding (PCCCing) schemes optimized under the assumption of no API at the decoder. Moreover, in the presence of strong API the proposed SCCCs allow to approach the Shannon limit (SL) more than any previously proposed turbo coding scheme.