International Journal of Engineering and Technology

This paper show the hardware implementation of the technique known as Compressive Sensing (CS). CS work develops in a sparse space (dense few). We used to obtain sparse signals Discrete Fourier Transform (DFT), Discrete Cosine Transform (DCT), and Discrete Wavelet Transform (DWT). We demostred that speech, electromyography and electrocardiogram signals can be reconstructed from few samples (projection). These projections were physically transmitted to a computer for reconstruction using L1 magic and demostrate that CS projections obtained were correct. The implementation CS showed relevant results, increase in speed and developing new communications. The implementation was performed in a Field Programmable Gate Array (FPGA) SPARTAN 3E and 18F4550 microcontroller. The results showed that is possible the reconstruct signals, breaking Shannon and Nyquist’s theory. Furthermore, we concluded that the FPGA implementation is faster and allows higher compression rates than the microcontroller. Finally, we showed using minimization L1 that DWT, presents better results in sparse conversion.