Turbo AI, Part III: Facilitating Wireless Massive Access for Next Generation PRACH

The contention-based massive access belongs to one of the most important features of the next generation wireless communications, in which the number of wireless personal devices will significantly increase. Within this context, technical evolutions can be expected for the next generation Physical Random Access Channel (PRACH) jointly with the Machine Learning (ML) based techniques. In this paper, we propose a novel PRACH framework, which can systematically generate a large number of PRACH preamble patterns for supporting massive access and identifying individual users. Furthermore, this PRACH framework is capable of facilitating the iterative ML-based detection, referred to as PRACH-Turbo-AI. Numerical results show that the proposed concept can improve the resolution of the PRACH preambles detection especially for low Signal-to-Noise Ratio (SNR), and exhibit great flexibility to impact both the current standard and next generation wireless systems complementarily.