Computation of correlation and cumulant 4 under FPGA architecture
Résumé
Higher Order Statistics are used in digital signal
processing as a powerful analytical tool for the analysis of
signals and systems. These statistics are very useful in problems
where either non-Gaussianity or non-minimum phase. The
computation of the fourth-order cross moments from incoming
time-series data is an intensive process and requires parallel
processing techniques and fast computing systems in order to
follow the requirements of real-time processing. This paper
presents an FPGA based design for high-speed computation of
fourth-order cross moments. The proposed design is coded in
VHDL and functionally verified by implementing it on Xilinx
Virtex-5 FPGA. Simulations results are presented. The
proposed design operates at a maximum frequency of 375 MHz.