FPGA realization of Caputo and Grünwald-Letnikov operators

Abstract

This paper proposes a hardware platform implementation on FPGA for two fractional-order derivative operators. The Grünwald-Letnikov and Caputo definitions are realized for different fractional orders. The realization is based on non-uniform segmentation algorithm with a variable lookup table. A generic implementation for Grünwald-Letnikov is proposed and a 32 bit Fixed Point Booth multiplier radix-4 is used for Caputo implementation. Carry look-ahead adder, multi-operand adder and booth multiplier are used to improve the performance and other techniques for area and delay minimization have been employed. A comparison between the two presented architectures is introduced. The proposed designs have been simulated using Xilinx ISE and realized on FPGA Xilinx virtex-5 XC5VLX50T. The total area of 2515 look up tables is achieved for Caputo implementation, and maximum frequency of 54.11 MHz and 1498 slices are achieved for Grünwald-Letnikov architecture. © 2017 IEEE.

Authors

Tolba M.F., Abdelaty A.M., Said L.A., Elwakil A.S., Azar A.T., Madian A.H., Ounnas A., Radwan A.G.

Keywords

Caputo; FPGA; Fractional calculus; Grünwald-Letnikov; LUT; Non-uniform segmentation

Document Type

Confrence Paper

Source

2017 6th International Conference on Modern Circuits and Systems Technologies, MOCAST 2017, Art. No. 7937659, Doi: 10.1109/MOCAST.2017.7937659

Scopus Link

Comments are closed.