An Analysis of WS and PFF Algorithms

In order to analyze WS and PFF algorithms, mathematical models have been developed. The results obtained from the analysis in this paper are T_min and T_max of the parameter T for the two algorithms respectively.
The input parameters for the models were obtained from the address traces of three programs. The results of the models were validated by simulation.
WS algorithm has better performance. It adapts to dynamic changes in program behavior during execution. Therefore its performance is expected to be less dependent on prior knowledge of program behavior. The parameter T can vary in a wider range. Therefore the value of T can be easily selected to make the algorithm work well.