Wearable electronics capable of recording and transmitting biosignals can provide convenient and pervasive health monitoring. A typical EEG recording produces large amount of data. Conventional compression methods cannot compress date below Nyquist rate, thus resulting in large amount of data even after compression. This needs large storage and hence long transmission time. Compressed sensing has proposed solution to this problem and given a way to compress data below Nyquist rate. In this paper, double temporal sparsity based reconstruction algorithm has been applied for the recovery of compressively sampled EEG data. The results are further improved by modifying the double temporal sparsity based reconstruction algorithm using schattern-p norm along with decorrelation transformation of EEG data before processing. The proposed modified double temporal sparsity based reconstruction algorithm out-perform block sparse bayesian learning and Rackness based compressed sensing algorithms in terms of SNDR and NMSE. Simulation results further show that the proposed algorithm has better convergence rate and less execution time.

Compress sensing has advantage of simultaneous sensing and data compression. In recent years block compressive sensing (BCS) is celebrated because of its low encoding complexity. In this paper FOCUSS reconstruction algorithm based on BCS frame work is proposed utilizing iterative re-weighted l ₁ norm minimization. In proposed approach the required sparsity was attained using l ₁ norm minimization and 2D-Median filter is implemented as smoothing operator for removing blocking artifacts. The performance of reconstruction algorithm is verified on different data sets and results obtained from analysis reveal that the suggested approach for image reconstruction offers significant compression performance in contrast to existing techniques.

The application of spread-spectrum signals (arbitrary pulse width and position (APWP) sequences) in air-coupled resonant ultrasound spectroscopy is studied. It was hypothesized that spread-spectrum signal optimization should be based on te signal to noise ratio (SNR). Six APWP signal optimization criteria were proposed for this purpose. Experimental measurements were conducted using a thin polycarbonate sample using two standard spread-spectrum signals, linear and nonlinear frequency modulation, together with six optimized APWP signals. It was found that the performance of APWP signals derived from linear frequency modulation was better. The two best performing optimization criteria are SNR improvement on a linear scale with the SNR as an additional weight and energy improvement on a dB scale. The influence of spectral coverage on measurement errors was evaluated. It was found that it is sufficient to cover the sample resonance peak and the valley. The lowest error rates for density, 3%, and for thickness, 3.5%, were achieved when the upper valley was covered. For velocity, the best result, 5%, was achieved when the lower valley was covered. The lowest error rate for attenuation, 3.8%, was achieved in the case when both valleys were covered. Yet no significant performance degradation was noted when a whole −30 dB passband was covered.

Interfering reflections occur due to air impedance mismatch from sample and transducer in non-contact resonance ultrasound spectroscopy (NC-RUS) measurements. The transducer is in close vicinity to the sample in order to ensure low attenuation in the air and flat wavefront. Therefore, interfering reflections are at 30-100 μs distance from the useful signal. Pulse excitation is used to avoid the overlap that can distort the measurement. A short pulse is required to ensure a broadband excitation, but such a pulse is low energy. Air-to-sample impedance mismatch further reduces signal energy and SNR. Therefore, very high voltage excitation and averaging is required. High voltage excitation increases nonlinearity effect in air, averaging increases measurement time. Long spread spectrum (SS) signals can achieve high SNR, avoid distortion in air, and expand the bandwidth available for analysis. However, long SS signals will overlap. Here it was demonstrated that the pulse compression (PuC) using matched filter is not suitable: the correlation sidelobes overlap with interfering reflections and bias errors of NC-RUS measurements increase. Solution proposed for PuC uses inverse filter. Such compressed signal corresponds to ideal pulse excitation (uniform spectrum) and has no sidelobes. If the spectral content of the SS needs to be preserved, then the PuC result can be gated and decompressed back. It was demonstrated that errors can be significantly reduced in such a case. For example, 0.008% sample density estimation bias error is achieved in case of 350 μs long excitation compared to 2.42% when using matched filter or 2.1% in case of mismatched filter, a 300 times reduction.