• DocumentCode
    1127574
  • Title

    On Optimal Outage in Relay Channels With General Fading Distributions

  • Author

    Atia, George ; Sharif, Masoud ; Saligrama, Venkatesh

  • Author_Institution
    Boston Univ., Boston
  • Volume
    53
  • Issue
    10
  • fYear
    2007
  • Firstpage
    3786
  • Lastpage
    3797
  • Abstract
    This correspondence deals with the outage capacity of relay networks in the low signal-to-noise ratio (SNR) regime. This work is motivated by the fact that in relay channels, unlike multi-antenna point-to-point links, the transmitters, i.e., source and relays, may not be co-located and therefore their channel statistics can be quite different. It has been recently shown that bursty amplify and forward (BAF) is outage optimal when all the links have a Rayleigh distribution. In this correspondence, it is shown that BAF is in fact outage optimal for a wide class of independent channels with smooth distribution functions. Optimality of BAF is further generalized to a special class of dependent channels, namely, the case where we opportunistically use only the best relay (out of N relays). It turns out that relative to the strategy where all the N relays are used, this opportunistic best relay strategy uses significantly smaller average power (independent of ) while suffering negligible increase in outage. This holds out potential for significant gains in an ad hoc network scenario where minimizing interference to possibly other users as well as conserving power are important considerations.
  • Keywords
    Rayleigh channels; ad hoc networks; channel capacity; statistical distributions; Rayleigh distribution; channel statistics; fading distribution; multiantenna point-to-point link; optimal outage; relay channel; Decoding; Digital relays; Distribution functions; Diversity methods; Engineering profession; Fading; Rayleigh channels; Signal to noise ratio; Statistical distributions; Transmitters; Cooperation; diversity; fading channels; outage capacity; relay channel;
  • fLanguage
    English
  • Journal_Title
    Information Theory, IEEE Transactions on
  • Publisher
    ieee
  • ISSN
    0018-9448
  • Type

    jour

  • DOI
    10.1109/TIT.2007.904965
  • Filename
    4305384