• DocumentCode
    1506486
  • Title

    Characteristics of Fast Physical Random Bit Generation Using Chaotic Semiconductor Lasers

  • Author

    Hirano, Kunihito ; Amano, Kazuya ; Uchida, Atsushi ; Naito, Sunao ; Inoue, Masaki ; Yoshimori, Shigeru ; Yoshimura, Kazuyuki ; Davis, Peter

  • Author_Institution
    Dept. of Electron. & Comput. Syst., Takushoku Univ., Hachioji, Japan
  • Volume
    45
  • Issue
    11
  • fYear
    2009
  • Firstpage
    1367
  • Lastpage
    1379
  • Abstract
    We investigate the characteristics of fast random bit generation using chaotic semiconductor lasers. The optical amplitudes of two lasers with chaotic oscillations induced by optical feedback are each sampled at a fixed rate to extract binary bit sequences which are then combined by an exclusive-OR operation to obtain a single random bit sequence. Bit sequences generated at rate of 1 Giga bit per second are verified to pass statistical tests of randomness. We describe the dependence of randomness on laser parameters, in particular the injection current, the external cavity length and the feedback strength. The results provide clear empirical guidelines for tuning the chaotic laser parameters to achieve random bit sequences. This study shows that chaotic laser devices can be fast and reliable sources of physical entropy for computing and communication applications.
  • Keywords
    laser feedback; laser tuning; optical chaos; random number generation; random sequences; semiconductor lasers; binary bit sequences; chaotic oscillations; chaotic semiconductor lasers; external cavity length; fast random bit generation; injection current; optical feedback; statistical randomness. tests; tuning; Chaos; Chaotic communication; Character generation; Guidelines; Laser feedback; Laser theory; Laser tuning; Optical feedback; Semiconductor lasers; Testing; Chaos; information technology; random number generation; semiconductor laser;
  • fLanguage
    English
  • Journal_Title
    Quantum Electronics, IEEE Journal of
  • Publisher
    ieee
  • ISSN
    0018-9197
  • Type

    jour

  • DOI
    10.1109/JQE.2009.2031310
  • Filename
    5291985