One of the important problems in the design of WOLF is how to find an efficient and easy-to-implement method that can separate active data from inactive data and put them into different buffers accordingly.
