Reducing switching activity in FIR filters by reordering the coefficients through the use of improved heuristic algorithm

The main goal of this paper is the switching activity reduction in FIR filters after the reordering of coefficients from both a heuristic algorithm named Anedma and its improved version. Due to the characteristics of these filters, which involve multiplications of input value with suitable coefficients, the best reordering of these operations can contribute for the reduction of the switching activity, what leads to the minimization of power consumption. The filters were synthesized using SYNOPSYS Design Compiler for the XFAB 180 nm technology, and the power estimation was realized at logic level with PrimeTime tools for sinusoidal vectors. The results show that it is possible to achieve average reduction in switching power of up to 13.18% after reordering the coefficients using the Anedma algorithm. By the results from its improved version it is also possible to notice that the Anedma algorithm is near to the optimal solution in terms of coefficients reordering.