WiSeR TinyOS Signal Processing Package

Author: Rajat Khanda(rkkhanda@mail.uh.edu)
WiSeR Group,University of Houston

1. Introduction:

While using memory and power constraint devices like moteiv for sensor data acquisition and data transfer using radio, it's not a novel idea to transfer all the data gathered after sampling. So we need to process the data to reduce the size of it. Thresholding, quantization, wavelet transform and compression help a lot to achieve this goal. The TinyOS signal Processing Package provides implementation of signal processing modules.

1.1 The signal processing package contains the following modules:

• Thresholding
• Quantization
• Wavelet Transformation
• Runlength Encoding
• LZ77 encoding

Two more files are provided for writing data into the Flash memory and reading from Flash using interface BlockRead and BlockWrite.

2. Platform Information :

Modules are compatible with TinyOS 2.0 and implemented using NESC.

3. What the modules will do?

• Thresholding module clips the input data value to zero below a certain threshold.
• Quantizaion module can be used to quantize the samples and to find the quantization index.
• Wavelet module implements Cohen-Daubechies-Feauveau transfrom.
• LZ77 and Runlength encoding are used for compression.

4. How to use the modules?

Download the source code from attachment below and wire up the respective component in your configuration file to use the interfaces provided by the modules.For svn checkout click here.

5. References

http://www.owlnet.rice.edu/~jryans/wt_tinyos.html
http://cvs.cens.ucla.edu/viewcvs/viewcvs.cgi/tos-contrib/WaveletCodec/tos/lib
http://en.wikipedia.org/wiki/LZ77