Least-Squares Design of Variable Maximally Linear FIR Differentiators
Jong-Jy Shyu, Soo-Chang Pei, Yun-Da Huang · IEEE Transactions on Signal Processing · 2009
In this correspondence, the design of variable maximally linear (VML) FIR differentiators by least-squares method is proposed. The VML differentiator is designed such that its maximally flat frequency is adjustable. The used transfer function is formulated in a variable frame, and the Farrow structure can be applied for the implementation of VML FIR differentiators. For comparison, the non-variable optimal design of maximally linear FIR differentiators is also proposed. The experiments show that the variable design indeed possesses the flat characteristics as the independent optimal design.