eduzhai > Applied Sciences > Engineering >

Block Deep Neural Network-Based Signal Detector for Generalized Spatial Modulation

  • king
  • (0) Download
  • 20210506
  • Save

... pages left unread,continue reading

Document pages: 5 pages

Abstract: Generalized Spatial Modulation (GSM) is being considered for high capacityand energy-efficient networks of the future. However, signal detection due tointer channel interference among the active antennas is a challenge in GSMsystems and is the focus of this letter. Specifically, we explore thefeasibility of using deep neural networks (DNN) for signal detection in GSM. Inparticular, we propose a block DNN (B-DNN) based architecture, where the activeantennas and their transmitted constellation symbols are detected by smallersub-DNNs. After $N$-ordinary DNN detection, the Euclidean distance-based softconstellation algorithm is implemented. The proposed B-DNN detector achieves aBER performance that is superior to traditional block zero-forcing (B-ZF) andblock minimum mean-squared error (B-MMSE) detection schemes and similar to thatof classical maximum likelihood (ML) detector. Further, the proposed methodrequires less computation time and is more accurate than alternativeconventional numerical methods.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×