Myopic policy for opportunistic access in cognitive radio networks by exploiting primary user feedbacks