A Probability Based Modulation and Its Application in Ultra Reliable Low Latency Communications in 5G

Abstract

The optimal modulations when the bits are not equally distributed between occurrence of 0 and 1 are studied. According to their probability distribution, the constellation points in our scheme are shaped via the Huffman coding algorithm and we call it a probability-based modulation in this paper. Mathematical analysis proves that, it can increase the minimum Euclidean distance (MED) of the conventional modulation by above 58\% and improves the performance significantly. In addition, its demodulation algorithm is as simple as that of conventional ones thanks to additional design principles proposed in this paper. Consequently, it can be a good candidate modulation for the non-uniformly distributed bits. Based on this observation, we adopt it in the hybrid automatic retransmission (HARQ) feedback channel for ultra reliable low latency communications (URLLC) in 5G. As one can see that, if 0 and 1 represent acknowledge (ACK) and non-acknowledge (NACK) message, respectively, the payload bits in this channel will be highly unevenly distributed between 0 and 1 due to the high reliability of this service. By use of the proposed modulation, a code division multiplexing (CDM) and repetition coding based design is proposed. Both theoretical analysis and simulation results show that our modulation based method outperforms the conventional modulation based one in terms of performance and power efficiency.