• DocumentCode
    265946
  • Title

    Increasing the throughput of HARQ via multi-packet transmission

  • Author

    El Hamss, Aata ; Szczecinski, Leszek ; Piantanida, Pablo

  • Author_Institution
    INRS-EMT, Montreal, QC, Canada
  • fYear
    2014
  • fDate
    8-12 Dec. 2014
  • Firstpage
    1485
  • Lastpage
    1491
  • Abstract
    In this work we consider incremental redundancy hybrid ARQ (HARQ) transmission of data packets over independent block-fading channels. In HARQ, the transmitter, upon request (NACK), resends the "NACKed" packet in the next available block. Conventionally, the original and retransmitted packets occupy one block, each. In our approach, we allow the transmitter to share the resources between the new and the retransmitted packets. To implement such a multi-packet transmission we consider the superposition coding and the timesharing modulation as possible encoding schemes for simultaneous transmission of both packets. Targeting the maximization of the throughput, we cast the problem into the framework of Markov decision process (MDP), where the decisions about the optimal transmission strategy are taken using information about the state of the receiver in the previous transmissions. Numerical examples obtained in a Rayleigh block-fading channel show that proposed transmission strategy provides notable gains over a conventional HARQ.
  • Keywords
    Markov processes; Rayleigh channels; automatic repeat request; data communication; decision theory; modulation coding; radio receivers; radio transmitters; HARQ transmission; MDP; Markov decision process; NACKed packet; Rayleigh block-fading channel; automatic repeat request; data packets; encoding schemes; hybrid ARQ transmission; multipacket transmission; superposition coding; time sharing modulation; transmission strategy; Decoding; Encoding; Optimization; Radiation detectors; Receivers; Throughput; Transmitters;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Global Communications Conference (GLOBECOM), 2014 IEEE
  • Conference_Location
    Austin, TX
  • Type

    conf

  • DOI
    10.1109/GLOCOM.2014.7037018
  • Filename
    7037018