1: \begin{abstract}

2: In this work, we propose a type of $G_N$-coset codes for parallel decoding.

3: The parallel decoder exploits two equivalent decoding graphs of $G_N$-coset codes.

4: For each decoding graph, the inner code part is composed of independent component codes to be decoded in parallel.

5: The extrinsic information of the code bits is obtained and iteratively exchanged between the two graphs until convergence.

6: Accordingly, we explore a heuristic and flexible code construction method (information set selection) for various information lengths and coding rates.

7: Compared to the previous successive cancellation algorithm, the parallel decoder avoids the \emph{serial} outer code processing and enjoys a higher degree of parallelism.

8: Furthermore, a flexible trade-off between performance and decoding latency can be achieved with three types of component decoders.

9: Simulation results demonstrate that the proposed encoder-decoder framework achieves comparable error correction performance to polar codes with a much lower decoding latency.

10: \end{abstract}

11: