@inproceedings{
  author = {M. Hummert and N. Bulk and C. Bockelmann and D. W\"{u}bben and A. Dekorsy},
  year = {2024},
  month = {Mar},
  title = {Improving NOMA Performance by Application of Autoencoders and Equidistant Power Allocation},
  URL = {https://wsa2024.org/},
  address={Dresden, Germany},
  abstract={Non-orthogonal multiple access (NOMA) has beenintroduced as a promising scheme to allow for superpositionof signals, such as the transmission of multiple services in thesame resource block (time and frequency). In this paper, wepropose the application of Deep Learning (DL) in an Autoencoder(AE) framework for simultaneous usage in multi-service NOMAtransmission. While classical NOMA simultaneously incorporateslocally separated user equipments (UEs), we focus on the simul-taneous transmission of services within a single UE. Our schemeachieves improved performance compared to classical NOMAschemes and is capable of performing optimal estimation of thesuperimposed transmit signals. The scheme utilizes an equidistantpower allocation scheme. The results show the potential of usingDL to enhance the performance of NOMA systems and improvetheir adaptability and flexibility to different scenarios.},
  booktitle={27th InternationalWorkshop on Smart Antennas (WSA)}
}