The impact of hardware implementation on the performance of spectrum sensing algorithms

The problem of efficient hardware implementation of a reliable spectrum sensing method is crucial for the practical realization of the cognitive radio terminal. This paper analyses the influence of various hardware implementation issues on the performance of the sequential energy-detection-based spectrum sensing algorithm. In particular, the necessity of noise inclusion in noise variance estimation is discussed. Furthermore, based on the experiments conducted with the USRP N210 board used for the detection of a wireless microphone, the impact of numerous unwanted phenomena on the reliability of the spectrum sensing algorithm is also presented. It has been stated that such effects can influence not only noise-power dependent methods, but also the algorithms of other classes.