A novel and fast way to decode VLEC

Abstract

Joint Source Channel Coding (JSCC) is an emerging area of Information Theory. It jointly optimises source code and channel code both. Variable Length Error Correcting Codes (VLECs) are the tools of JSCC and they are very important as there are a lot of applications of VLECs in the transmission of multimedia files. There are several decoding algorithms of VLECs, one of them is Maximum Likelihood (ML) decoding. This uses trellis/tree structure to decode the received bitstream. The problem with the trellis approach is its complexity which increases exponentially with the length of the received vector. This paper proposes a rather novel and fast tabular decoding method for ML decoding of VLECs. In place of the trellis or tree structure, if tabular method is used, the complexity of the decoding algorithm reduces a lot.