Diversity techniques for spectrum sensing in fading environments

A key feature of a cognitive radio (CR) is a reliable spectrum sensing technique which enables the CR to detect the so-called white spaces in the frequency band. This allows opportunistic access of the unlicensed (secondary) users to these white spaces without causing undue interference to licensed (primary) users. In many scenarios the CR may operate in a multipath fading environment where spectrum sensing must cope with the fading effects of the unknown primary signal. In this paper we first study the effects of multipath fading of the performance of the autocorrelation-based spectrum sensing algorithm. The results show that Rayleigh fading causes significant degradation in the detection and false alarm. This motivates us to investigate three diversity combining techniques, namely equal gain combining, selective combining, and equal gain correlation combining. For Rayleigh fading channels we evaluate the performance of these three techniques through simulation. The results show that for detection probabilities of interest (e.g., > .9), a system with a four-branch diversity achieves an SNR gain of more than 5 dB over the no-diversity system that uses the same number of received signal samples.