A new discrete cosine transform-based energy detection scheme investigation using bartlett periodogram method in cognitive radio spectrum sensing

Cognitive Radio (CR) is a promising technique through which the efficiency of utilization of the electromagnetic spectrum can be increased in various ways. CR lets unlicensed secondary users (SUs) utilize the licensed spectrum during the period of no transmission where the primary users (PUs) are in an idle state; this will improve the spectrum utilization to deal with spectrum scarcity. By using spectrum sensing in cognitive radio, the unused spectrum can be sensed, exploited and used to fill the lack of bands in the new applications. The current energy detectors which are used to reduce the scarcity of the spectrum suffer from weak performance due to high noise variance, low signal resolution and low signal-to-noise ratio, especially when it is working in the frequency domain. This thesis focuses on the Bartlett periodogram method using the discrete cosine transform (DCT) instead of previous methods which used the discrete Fourier transform (DFT) to get the power spectrum density that is used to detect the primary user by comparing with the predefined threshold. Digital Video Broadcast-Terrestrial (DVB-T) signals are used as an application example to analyze and assess the proposed spectrum sensing algorithm in the frequency domain in the AWGN channel. The accuracy of the proposed analysis is confirmed by using Monte Carlo trials. The results show an accurate performance analysis of the Bartlett periodogram based on DCT, reducing the noise variance without decreasing the signal resolution compared with the Bartlett periodogram based on FFT. In this proposed method the average noise variance in the DCT Bartlett’s periodogram in all the scenarios is 0.35385 for the QPSK and 0.3478 for 16-QAM, while the average noise variance in the FFT Bartlett’s periodogram in all the scenarios is 11686.5 for the QPSK and 5841.37 for 16-QAM. This is because using the DCT transform yields better results compared to the FFT transform as shown in the simulation results. This is mainly due to DCT being a real transform possessing an energy compaction property. The leakage effect is not there for DCT as compared to DFT. The better spectrum sensing algorithm would require some trade-off between probability of detection (PD) and the probability of false alarm (PFA) to obtain good accuracy with low noise variance.

Preview Text FK 2016 158 - IR.pdf Download (2MB) | Preview

Actions (login required)

View Item