Decoding for wireless super-dense networks and its finite-length analysis for practical applications

One of the most important solutions for spectrum shortage, expected by 2020, is wireless networks with super density, which is also expected to provide significant economic impacts. This paper proposes decoding scheme with practical analysis for massive number of users in super-dense wireless networks involving two multiway relays (SD-MWMR). The practicality is viewed from the point of finite-length analysis, where the frame length is set very short to guarantee low latency networks, e.g., below 1 milisecond as suggested by 5G requirement. Due to the nature of finite block length, the probability of appearing stopping sets is high. This paper derives the probability of stopping sets (with regular and irregular degree distributions) in SD-MWMR to predict the theoretical performances. Extrinsic information transfer (EXIT) charts are also used to confirm the decoding behavior of SD-MWMR networks. The results confirm that the theoretical analysis and simulation results match each other indicating that two relays can detect more users or devices and at the same time reduce significantly the error-floor due to less occurrence probability of stopping sets in practice with finite-length setting.