Combination of short term and long term database for cognitive radio resource management

We propose a method that uses long term information on the use of primary channels to select the most auspicious ones to be sensed and exploited by cognitive radios at the requesting time. These channels are investigated in more detail over the short term. Sensing results are stored in the short term database and used to predict which channels are best for data transmission. The method makes the operation of cognitive radios more reliable and efficient in terms of delay and throughput, and decreases collisions with primary users.