The demand for the radio frequency (RF) is increasing to support the user's needs in wireless communication. RF spectrum is a scarce resource and requires efficient utilization. The most important methods for solving this problem are to let wireless users to share wide access spectrum. Due to the limitation of access spectrum and helpless uses of spectrum, a new idea is needed that is known as a cognitive radio. Cognitive radio has been proposed to promote the efficient use of spectrum. In this method, secondary users use the spectrum if a part of spectrum is not used by the primary users. To detect the presence of the primary users, the spectrum sensing is one of the fundamental enabling technologies for cognitive radio communication. This ensures the efficient utilization of the spectrum. According to cognitive radio, the secondary users might not have sufficient data to detect the primary users. There are some methods for signal detection such as cyclostationary feature detector, energy detector, autocorrelation detector and matched filter detector. The aim of this research is to propose a new method for detection of LTE (long term evolution). One of the properties of LTE is preamble that it uses CAZAC (Constant Amplitude Zero Autocorrelation Sequence). CAZAC, is one of the polyphase code, has a lot of applications in channel estimation and equalization and it shows periodic correlation properties. The final aim of this research is to use of this sequence to detect LTE as a primary user and to improve detecting of primary users in cognitive radio. In this research, CFAR (constant false alarm rate) methods are used in noise power estimation such as CA-CFAR and determine threshold based on noise power so that the effect of noise power is decreased in the proposed method. Finally, the proposed method is compared to other methods such as energy detector and autocorrelation detector and it is shown that the proposed method has better performance than energy detector. In order to decrease the computional complexing, it is shown that sign bit could be used in detection. Moreover, to decrease the effect of interference in desirable signal detection, the use of other CFAR methods such as CML and FCME are possible. Keywords: cognitive radio, spectrum sensing, cross-correlation, CAZAC.
