Deep Learning Detectors with Pilot Interval Reduction in Uplink Non Orthogonal Multiple Access
Ahmet Emir, Ferdi Kara, Hakan Kaya
proceeding (Uluslararası, Hakemli) | Yazar(lar): Ahmet Emir, Ferdi Kara, Hakan Kaya | Yıl: 2020 | Dergi/Kurum: 2020 28TH SIGNAL PROCESSING AND COMMUNICATIONS APPLICATIONS CONFERENCE (SIU) | İndeks: WOS.ISTP | Anahtar Kelimeler: Non-orthogonal multiple Access (NOMA), deep learning, Pilot Interval Reduction — Özet: Non-O
Non-Orthogonal Multiple Access (NOMA) has higher spectral efficiency than orthogonal multiple access (OMA) techniques. In uplink communication systems that the channel is not known at the receiver, pilot signals sent from each user in different time intervals have reduced the spectral efficiency of NOMA. In this study, in the uplink communication system, DLdeep learning based detectors which channel state information is not known at the receiver have been researched. It is aimed to maintain the spectral efficiency of NOMA by sending a single pilot from users, thus reducing the time interval in the DL detectors.