Low Complexity Decoding for Spinal Codes: Sliding Feedback Decoding

Siyue Xu,Shaohua Wu,Junxin Luo,Jian Jiao,Qinyu Zhang
DOI: https://doi.org/10.1109/vtcfall.2017.8287919
2017-01-01
Abstract:As a type of newly invented rateless codes, Spinal codes are characterized by capacity achieving over both additive white Gaussian noise (AWGN) and binary symmetric channel (BSC) with short message length and pseudo-random like code-words. For the emerging ultra-reliable low-latency communication (URLLC) scenarios such as information exchanging between self-drive cars, Spinal codes hold great prospects. However, the high decoding complexity of Spinal codes remains a bottleneck for its practical applications. In this work, a novel low complexity decoding algorithm named sliding feedback decoding (SFD) for Spinal codes is proposed. By 'sliding', the decoding tree is layered by a sliding window. By 'feedback decoding', the optimal parent node decision for each layer located by the sliding window is made by the feedback from the best leaf node in the located layer. And the final decoding path is composed of all the optimal parent nodes selected layer by layer. The complexity of the proposed algorithm is analyzed theoretically, and the results show that it is lower than the complexity of other algorithm. Extensive simulations are carried out to verify the effectiveness of the proposed algorithm. Compared with the bubble decoder and the forward stack decoding (FSD) proposed in the literature, SFD can significantly reduce the decoding complexity without any harming to the rate performance.
What problem does this paper attempt to address?