On the Concatenation of Superposition and Polar Codes

Concatenated codes are currently being actively investigated for the next-generation wireless communication systems due to their flexibility and ability to be used in various scenarios. The serial type of concatenation provides good error-correcting properties with a potential reduction in complexity when an iterative concatenated decoder is used at the receiver side. One of the most effective approaches nowadays is based on integrating an error-correcting code as an outer component and an error-reducing code as an inner component of concatenated construction. There are several approaches to construct a code that has an error-reducing property. Sparse regression codes (SPARC), which are currently being actively researched, may be a promising candidate. These codes exhibit a property to reduce the number of errors in additive white Gaussian noise (AWGN) channel, but they have a high decoding complexity. We propose concatenated scheme with outer polar code and convolutional-based superposition inner code. The latter code is a superposition of BPSK-modulated convolutional codes with successive interference cancellation (SIC) decoder. The use of such a decoder for inner component renders the proposed approach practically feasible.