• DocumentCode
    2934712
  • Title

    Physical layer key generation methods for arbitrary fading channels

  • Author

    Sharma, Rajesh K. ; Wallace, Jon W.

  • Author_Institution
    Sch. of Eng. & Sci., Jacobs Univ. Bremen, Bremen, Germany
  • fYear
    2011
  • fDate
    3-8 July 2011
  • Firstpage
    1368
  • Lastpage
    1371
  • Abstract
    Two methods are developed that can be applied to generate secret keys automatically from fading propagation channels that are reciprocal but have arbitrary (i.e. non Gaussian) statistics. Such methods may be necessary for physical-layer key generation in cases where the line-of-sight (LOS) component produces Rician channel statistics, not only because channel quantization based on Gaussian statistics will not provide equally probable symbols, but also because the symbol error rate (SER) and efficiency analysis based on Gaussian channels does not reflect true performance. An improved channel quantization method compared to is developed, where the empirical cumulative distribution function (cdf) of the channel is used directly to ensure equal probability of the key symbols. The results show that LOS channels can have slightly better SER performance than strictly Gaussian channels, especially at low SNR. Second, the idea of positional coding is developed, where a secret key can be transmitted by dividing empirical channel observations into multiple codewords and conveying a secret message from Alice to Bob in the sequence of channels fed forward from Alice to Bob. Analysis of the method illustrates that key mismatch rate can be made arbitrarily low by properly selecting the codeword length.
  • Keywords
    Gaussian channels; Gaussian processes; Rician channels; encoding; error statistics; quantisation (signal); statistical distributions; Gaussian channels; Gaussian statistics; LOS channels; Rician channel statistics; channel quantization method; codeword length; empirical cumulative distribution function; fading propagation channels; physical layer key generation methods; positional coding; symbol error rate; Channel estimation; Decoding; Encoding; Fading; MIMO; Quantization; Signal to noise ratio;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Antennas and Propagation (APSURSI), 2011 IEEE International Symposium on
  • Conference_Location
    Spokane, WA
  • ISSN
    1522-3965
  • Print_ISBN
    978-1-4244-9562-7
  • Type

    conf

  • DOI
    10.1109/APS.2011.5996545
  • Filename
    5996545