eduzhai > Applied Sciences > Engineering >

Deep Learning Enabled Semantic Communication Systems

  • Save

... pages left unread,continue reading

Document pages: 13 pages

Abstract: Recently, deep learned enabled end-to-end (E2E) communication systems havebeen developed to merge all physical layer blocks in the traditionalcommunication systems, which make joint transceiver optimization possible.Powered by deep learning, natural language processing (NLP) has achieved greatsuccess in analyzing and understanding large amounts of language texts.Inspired by research results in both areas, we aim to providing a new view oncommunication systems from the semantic level. Particularly, we propose a deeplearning based semantic communication system, named DeepSC, for texttransmission. Based on the Transformer, the DeepSC aims at maximizing thesystem capacity and minimizing the semantic errors by recovering the meaning ofsentences, rather than bit- or symbol-errors in traditional communications.Moreover, transfer learning is used to ensure the DeepSC applicable todifferent communication environments and to accelerate the model trainingprocess. To justify the performance of semantic communications accurately, wealso initialize a new metric, named sentence similarity. Compared with thetraditional communication system without considering semantic informationexchange, the proposed DeepSC is more robust to channel variation and is ableto achieve better performance, especially in the low signal-to-noise (SNR)regime, as demonstrated by the extensive simulation results.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×