Performance analysis of periodical spectrum sensing for dynamic spectrum access networks

More Info
expand_more

Abstract

In Dynamic Spectrum Access Networks, detection of spectral opportunity can be viewed as a frequency scanning process leading to detection of legitimate users of radio channel. Continuous scanning results in complete detection but it becomes inefficient in terms of power consumption at each mobile node. This is more profound whem some frequency bands are rarely used by legitmate users. Therefore, it is necessary to study the scanning process which is periodically switched off according to some distribution to reduce power consumption. We analytically show that with the increasing number of scanning nodes thier average scanning-on time decrease while the efficiency approaches to the one node case. Based on the results we propose a scanning scheme for single and multiple node cases, that dynamically assign average scanning-on time to nodes when the number of nodes in the network is varying with time. At the scanning-off instance, depending on whether we allow transmission or not, we come across two cases which we also analyze: interfering and non-intefering.