Finite Length Non-binary Raptor Codes under Ordered Statistics Decoder
Fei Wang, Jian Jiao, Lianqin Li, Ke Zhang, Shaohua Wu, Qinyu Zhang · 2020
Raptor codes can approach the capacity of era-sure channel without accurate channel state information at the transmitter side, which is viewed as a potential channel coding approach to meet stringent requirements of ultra-reliable low latency communications (uRLLC) (block error rate (BLER) ≤ 10-5, and end-to-end latency ≤1 ms). This letter investigates a modified ordered statistics decoder (mOSD) algorithm for finite length non-binary Raptor code towards uRLLC. The upper bound of BLER for the non-binary Raptor code under OSD is derived, which can estimate the required block length under certain reliability requirement. Simulation results show that the BLER can be lower than 10-5in the finite length regime (-5, which is reduced by up to 50% of the average decoding complexity than the conventional OSD.