A hybrid Sphere Decoding for short polar codes using variable step size

In this paper, a hybrid stack Sphere Decoding (SD) for short polar codes based on radius extension and a novel path metric (PM) is proposed. The proposed SD scheme utilizes variable step size obtained from a novel PM, which is a modified PM from a previous work. The proposed method continuously increases the corresponding variable radius until a valid polar codeword is found. Moreover, by introducing a limited size stack, the proposed scheme can avoid node revisiting and achieve a trade-off between performance and complexity. Simulation results show that for short polar codes, compared with existing SD schemes that achieve maximum likelihood (ML) performance, the proposed scheme achieves more complexity reduction without compromising the ML optimality. (C) 2021 Elsevier B.V. All rights reserved.