This is a very active research area. It will shape the future for green frequency utilization using cognitive radio and network techniques. I would highly recommend such area for further research.

Good Luck

Thanks Mr. Ahmed sir.

Spectrum sensing is the backbone of CR networks. Efficient SS ensures efficient and interference free transmission. There are many SS techniques in literature. You may find the below mentioned article interesting and a starting point for your research.

Thanks Shahryar sir

in today's wireless communications, because of bandwidth demand these area is becoming very popular. Along with spectrum sensing in cognitive radio, you may focus on TV white space sensing. In fact spectrum sensing and communication are the main features of cognitive radio.  

Over the years I have been working on cooperative spectrum sensing and related security issues. Spectrum Sensing has many branches like energy detection, cyclo-stationary feature detection, matched filter, and eigen vector detection. Please remember that the goal of spectrum sensing is to reliably identify whether the spectrum is available for use. There are other popular alternatives, like the geo-location spectrum data base, where primary reports its usage. However, this is considered not very efficient in terns of spectrum utilization although it assures primary transmitter and receiver protection. The main challenge in spectrum sensing, is the security aspect, because a stand alone CR device's sensing result is not enough (due to typical wireless channel impairments). Hence multiple geographically proximate node's sensing results are fused. Now the spectrum decision maker is totally dependent on spectrum sensing results on individual CR devices which could be manipulated. Hence, many primary spectrum owners are still not convinced unless proper security measures are devised. At the end of the day, if the research community provides security along with spectrum sensing, then it will be a big in future, due to the spatial diversity it is able to provide. Spectrum sharing is another alternative option, but it is largely a policy issue rather than an elaborate research issue.

CRN is only based upon spectrum sensing, u cannot proceed further if spectrum is not sensed. How ever you require good knowledge of signals, signal processing etc. to succeed in this topic