<p>Chaotic systems with complex nonlinear behavior are one of the best candidates for random sequence resources to satisfy the increasing demand for security in cryptosystems and communications. However, most existing chaotic systems suffer from low throughput and insufficient randomness in hardware implementation. In this context, we first ascertain the low fluctuation in MSBs and LSBs, which is caused by fixed-point design of chaotic systems deployed on digital circuits. Then we propose the Modulated Rotated Tent map (MRTM) by using a novel method of switching the high and low fluctuation bit positions which prevents the incessant multiplication between low fluctuation bits. This method can significantly improve chaotic behavior and can be easily done in field-programmable gate arrays (FPGAs). Finally, a high–efficiency pipelined design of an FPGA-based PRNG is launched, which utilizes four cross-coupled channels to ensure both throughput and randomness. The uncorrelated chaotic trajectories drive two control registers to adjust the rotation degree between channels, which ensures a high unpredictability of the outputs to pass the famous test suites such as TestU01, NIST SP800-22, and Dieharder. Meanwhile, the system reveals merely 0.5 percent hardware resource utilization of the targeting FPGA while the timing report shows the system can remain stable at a clock up to 371 MHz in a Kintex-7 FPGA with a maximum throughput of 22.26 Gbps, which outperforms the state-of-the-art.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Designing high-efficiency pseudorandom number generators via the modulated rotated tent map

  • Yinzhe Liu,
  • Juan Du,
  • Jungang Miao,
  • Shouliang Li,
  • Zhen Yang

摘要

Chaotic systems with complex nonlinear behavior are one of the best candidates for random sequence resources to satisfy the increasing demand for security in cryptosystems and communications. However, most existing chaotic systems suffer from low throughput and insufficient randomness in hardware implementation. In this context, we first ascertain the low fluctuation in MSBs and LSBs, which is caused by fixed-point design of chaotic systems deployed on digital circuits. Then we propose the Modulated Rotated Tent map (MRTM) by using a novel method of switching the high and low fluctuation bit positions which prevents the incessant multiplication between low fluctuation bits. This method can significantly improve chaotic behavior and can be easily done in field-programmable gate arrays (FPGAs). Finally, a high–efficiency pipelined design of an FPGA-based PRNG is launched, which utilizes four cross-coupled channels to ensure both throughput and randomness. The uncorrelated chaotic trajectories drive two control registers to adjust the rotation degree between channels, which ensures a high unpredictability of the outputs to pass the famous test suites such as TestU01, NIST SP800-22, and Dieharder. Meanwhile, the system reveals merely 0.5 percent hardware resource utilization of the targeting FPGA while the timing report shows the system can remain stable at a clock up to 371 MHz in a Kintex-7 FPGA with a maximum throughput of 22.26 Gbps, which outperforms the state-of-the-art.