Enhanced Detectability Using Multi-Cycle Cyclostationary Detector in Cognitive Radio

1.Introduction

In the present scenario where RF spectrum is fully occupied by various wireless operators and applications. To meet the futuristic demand of 5G wireless networking, 8K video, unmanned driving etc. the requirement of high data rate, low latency are required. Spectrum sensing scheme focuses to resolve the problem of increasing data rate demand. Different types of spectrum sensing methods are being proposed and implemented in a move to choose the most optimal detector in terms of simplicity and detector performance (Cabric, 2004; Kamil et al., 2009; Subhedar & Birajdar, 2011). In this paper Cyclostationary feature detector performance is evaluated and an approach is made to improve its detectability by boosting the pilot subcarriers, by averaging over several symbols and by applying Multi-cycle cyclostationary algorithm. The performance is evaluated in terms of performance metric curve. For reference analysis Energy detector is considered (Abdulsattar & Hussein, 2012; Digham, 2003; Qingchun & Qilian, 2007) to make the base for comaparison, to denote the percentage improve in detection performance of detector under different schemes.

Cyclostationary detection scheme is capable of detecting and identifying analog and digital modulated communication signals (Aparna & Jayasheela, 2012; Aparna & Jayasheela, 2013; Verma et al., n.d.). The basic characteristic of modulated signal is periodicity of its statistical parameters like mean and autocorrelation because its spectral pattern repeats after certain period. The key feature of this scheme highlights its feature of outperforming other schemes in condition of extreme SNR scenario and it possesses a realistic model for distinguishing differently modulated signals by method of spectral analysis. Spectral correlation density function (SCD) is the tool used for signal analysis in cyclostationary detector. The found autocorrelation pattern is used to find SCD of the signal by applying cyclic Wiener-Khinchin relation (Song, 2011). Time-smoothed and frequency-smoothed cyclic periodogram is used to estimate the presence of signal. In the literature this work has been investigated for finding appropriate detectibility (Sohn et al., 2009) where the cyclostationary detection for the OFDM signal has been made by boosting and averaging over symbols, this is an enhancement work to explore and improve the detectibility by using concept of multicycle cyclostationary detection. The motivation of this research work has been taken for comprising these schemes for finding the best of all these schemes. In this paper the signal sensing is made in the frequency domain by doing the periodogram analysis.