Hardware Implementation of Local Mean Decomposition

  • 賴 秉暄

Student thesis: Master's Thesis


In many domains of modern science signal analysis is very important An appropriate signal analysis method is essential Local Mean Decomposition (LMD) was proposed by Jonathan S Smith in 2005 It is an adaptive signal decomposition algorithm It is effective in processing not only non-stationary signal but also non-linear signal According to these decomposed signals we can get the information such as instantaneous frequency from its features In recent years LMD has been gradually applied in many domains In some aspects it needs real-time precise calculation in LMD processing However it takes time to fulfill the LMD operation Improving the calculation becomes an important issue In the past years the design of LMD had not been proposed in Very-Large-Scale Integration (VLSI) or Field Programmable Gate Array (FPGA) To promote the performance and save the resource we do research on LMD algorithm in detail to optimize it according to its character We propose the architecture based on VLSI in our design in this thesis With the optimization of every module we realize a low-cost low complexity and high performance architecture We choose different precisions based on the user choice Also the different phenomenon is that the different modules can be replaced so that the user can choose the most coincident requirement of the LMD signal decomposition
Date of Award2014 Jul 16
Original languageEnglish
SupervisorPei-Yin Chen (Supervisor)

Cite this