RF spektrum algılamasına dayalı rastgele sayı üreticisi: enerji dedektörü ve spektral korelasyon fonksiyonu yaklaşımı

Abstract

Wireless communication is open and more vulnerable against to authorized and unauthorized users due to the broadcasting nature of wireless radio propagation in contrast to wired networks where two devices connected to each other physically through cables. Thus, assuring a secure wireless radio communication is an important and mandatory task for the 5G and beyond wireless networks. In order to prevent the manipulation and to ensure the privacy of the information, secure cryptographic algorithms are necessary. The performance of the cryptographic algorithms heavily relies on the generation of random keys which are created from the seeds and these seeds must be random. By utilization of the random nature of the wireless spectrum, secure random keys can be produced. Therefore, in this study, spectrum sensing based random number generator (RNG) is proposed in order to detect the unknown received signal and extract the noise part of the signal by simply adopting the second order statistics of the cyclostationary process, spectral correlation function and the energy detector approaches. However, utilization of probability mass function output statistics is also introduced to distinguish the noise and unknown signal. A measurement setup is developed also considering line of sight conditions. Obtained noise statistics are used to generate the random bit streams and the results are fed into the NIST 800-22 test suite to show how well the performance of the spectrum sensing based random number generator. High quality random numbers are obtained which implies that spectrum sensing based RNG can provide secure data transfer directly without any other physical device.