Direction of Arrival (DoA) estimation algorithms like MUSIC, Capon and ESPRIT can be used for spectrum sensing in cognitive radio networks. DoA so can be used for target user so its estimates give idea of direction. infact modification of DoA estimation methods like adaptive thresh-holding can increase capacity as well. See this paper

On the performance of DoA estimation algorithms in cognitive radio networks: -A new approach in spectrum sensing

In DOA estimation you can locate the direction of one or two sources so when you locate the direction you can cognitive radio source 

Yes DoA can be used in spectrum sensing for the purpose of PU detection and localization. Joint estimation of primary users and secondary users is a challenging task in CRNs but it can certainly improve the performance of CRNs by exploiting the property of spatial spectrum sensing. You can read the following article for further details 

http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=6942206&sortType%3Dasc_p_Sequence%26filter%3DAND(p_IS_Number%3A4356907)

There is no relation between DoA and CR spectrum sensing. DoA indicates which direction primary user wave is coming. But sensing means from the receiving wave, we have to find the frequency /power of the transmitted wave. The general fact is wave is available in all the directions.

best wishes

R.Suresh Babu

DOA estimation can help to find better spectrum sensing and achieving the threshold which depends on the distance of transmitter as well as noise level. It should be noted that DOA estimation needs to have array of antenna or imaginary array by using different base stations information.

DOA can be like physical layer information but CR can add more on to it and validate incoming data..  It will be good if you deploy CR based DOA finding equipment's.  In IoT, today is to find route and validate given routes before start communicating....it is same in DOA as well