Discrete Wavelet Transform(dB4)

Hi, I want to know whether we can do Discrete wavelet transform (dB4) in MSP430F5529?

Your requirement are not very detailed, but probably yes.

Is any inbuilt function in C language to do the same?

A special feature of the C language is that it does not have any "inbuilt" functions.

There is a set of standardized (i.e. POSIX) libraries (C-Lib, math lib) which are designed for full-blown OS (Unix), but often only subsets (newlib, nanolib) are in use in the embedded field.

Special libraries for wavelet transformations are available, just try your favourite search engine. But I would carefully check if the MSP430 is a match for your performance and resource requirements. This MCU is designed for low-power and general-purpose applications, not for heavy number crunching. Consider a Cortex M4 with FPU instead (MSP432).

I would like to do it in MSP430 microcontroller.
If you have any links related to this plz Send it.
Thank you for your kind answer and time.

I have no specific links to libraries you couldn't find yourself.
The stuff I found in a superficial search was mostly C++, Java or scripting (Matlab, R). This would be hard to swallow for a MSP430 ...

TI has fixed point math libraries (www.ti.com/.../msp-iqmathlib), which cover the "math basics", but no transformations.
The CMSIS DSP lib (for Cortex M) has transformation functions, not sure if anything for wavelets. But that would be applicable for the MSP432.

Thanks......
Is it possible to do convolution in MSP430F5529???
if it is possible i can supply low pass filter and high pass filter constants manually and do convolution. Then down sampling we will get detailed and approximate coefficient.

Is it possible to do convolution in MSP430F5529???

Within the Flash (code) and RAM (data) size constraints, you can implement almost any algorithm. But more important are your requirements. How fast needs it to be done ? What a resolution (block size/input size) is required ?

As an example, FFT implementations are occasionally discussed here. FFTW, a renowned library for Unix/Linux/Windows, supports arbitrary input and FFT block sizes. Doing an audio signal analysis on 44.1kHz input data with a 1Hz spectral resolution is not uncommon. That would be hardly possible with a 16-bit MCU, having a few kilobytes of RAM.

In my project my signal is sampled in 180Hz. Is it possible??

I have not tried any wavelet transformation yet (mostly Fourier, i.e. FFT), and on Cortex M and PCs. So don't rely solely on my judgment.
But assuming similar resource and performance requirements, 180Hz seems feasible to me.
Finally, it is your responsibility to find an implementation of the DWT algorithm (or do it yourself), and confirm that the MCU resources (performance, RAM size, code size) are sufficient.
Such preinvestigations usually yield a prototype (proof-of-concept), and are a significant part of the engineering business.

Thanks
I am doing Mtech in Embedded system.Above things are related my main project Thank you for kind replies

I am doing Mtech in Embedded system.

I'm not so firm in the latest "formal" terminology, but I assume a educational background. Your description did not really exclude a commercial project.

Not sure if you are expected to your own implementation - probably not. Finding an appropriate library, understanding and using it (instead of re-inventing the wheel) is an equally wanted skill.

From experience, I would suggest to rather invest some more hours/days in investigation and recherche, before going in a wrong direction. Even a lot of high-budget commercial projects fail for this reason ...