2021/04/24 by Pavlos S. Bouzinis, Panagiotis D. Diamantoulakis, Bouzinis, Pavlos S. +3
Computer Science · Engineering · Mathematics · #Advanced Wireless Communication Technologies #Artificial Intelligence (cs.AI) #Benchmark (surveying) #Channel (broadcasting) #Computer network #Computer science #Cooperative Communication and Network Coding #Decoding methods #FOS: Computer and information sciences #Full-Duplex Wireless Communications #Information Theory (cs.IT) #Interference (communication) #Mathematics #Noma #Protocol (science) #Reduction (mathematics) #Single antenna interference cancellation #Telecommunications #Telecommunications link #Wireless #Wireless network #cs.AI #cs.IT #math.IT
paper · pdf · doi:10.48550/arxiv.2104.12005
published in arXiv (Cornell University) (Cornell University)
arxiv created 2021/04/24 · openalex publication_date 2021/04/24 · arxiv updated 2021/04/27 · openalex created_date 2022/07/25 · openalex updated_date 2026/08/05
As it has been discussed in the first part of this work, the utilization of\nadvanced multiple access protocols and the joint optimization of the\ncommunication and computing resources can facilitate the reduction of delay for\nwireless federated learning (WFL), which is of paramount importance for the\nefficient integration of WFL in the sixth generation of wireless networks (6G).\nTo this end, in this second part we introduce and optimize a novel\ncommunication protocol for WFL networks, that is based on non-orthogonal\nmultiple access (NOMA). More specifically, the Compute-then-Transmit NOMA\n(CT-NOMA) protocol is introduced, where users terminate concurrently the local\nmodel training and then simultaneously transmit the trained parameters to the\ncentral server. Moreover, two different detection schemes for the mitigation of\ninter-user interference in NOMA are considered and evaluated, which correspond\nto fixed and variable decoding order during the successive interference\ncancellation process. Furthermore, the computation and communication resources\nare jointly optimized for both considered schemes, with the aim to minimize the\ntotal delay during a WFL communication round. Finally, the simulation results\nverify the effectiveness of CT-NOMA in terms of delay reduction, compared to\nthe considered benchmark that is based on time-division multiple access.\n