Cyclostationarity Feature Detection to IFDMA Uplink System in Cognitive Radio Networks

摘要

Reliable and efficient spectrum sensing is the key function of cognitive radio (CR) which act as sec­ondary users (SU) and dynamically access vacant frequency bands. However, the operation of primary users (PU) should be protected and the interference from secondary users should be avoided. This function is fulfilled by detecting the existence of PUs' signal located in such frequency band. Cyclostationarity feature detection is the main spectrum sensing technique which utilizes modulated signal inherent cyclostationarity characteristics. Almost all modulated signals show cyclostationarity characteristics which are suitable for signal detection. Some specific signals (e.g. Interleaved Frequency Division Multiple Access (IFDMA) signal) regularly distributed in the frequency domain possess specific cyclostationarity distribution feature which can be utilized to construct a signal detector. The detection of IFDMA signal is critical for Ultra-Wideband (UWB) terminals that share the same band with the 4G mobile communications. In this paper, we propose a new cyclostationarity feature generation scheme which is based on signal inherent frequency distribution. In this scheme, the window coefficient is presented to locate the specific points of cyclostationarity distribution. Compared with traditional cyclostationary detection methods, the proposed one is sample and efficient. This paper gives an insight that signal inherent distribution information and signal cyclic periodogram (CP) distribution information are useful for the design of simple and efficient spectrum sensors.