A Compressed Sensing Analog-to-Information Converter With Edge-Triggered SAR ADC Core

Compressed Sensing EEG Measurement Technique with Normally Distributed Sampling Series

“…6, compares the reconstruction accuracy and the amount of wirelessly transmitted information as the uniform distribution of the conventional method and the normal distribution of the proposed method. Figure 6 shows that the reconstruction accuracy at M = 102 (σ = 1.2) is the almost same as that of the conventional method (M = 115 [10]). An example of the sampling series of the four-channel MUX created in this simulation is shown in Fig.…”

“…Finally, the energy efficiency of the ADC with the proposed sampling series is compared with the conventional one [10]. In designing an ADC circuit, the number of samples as well as f s inter val often affects the power consumption.…”

“…Nyquist sampling [10] 92 Conventional (Uniform Distribution) [10] 12.2 Proposal (Normal Distribution) * 10.8 * The FoM of the proposed method is estimated for the same ADC based on in [10].…”

“…The sampling period of the narrowest sampling interval must be increased to reduce the maximum sampling frequency. During the sampling of a continuous uniform distribution with a widened sampling interval defined in the conventional method, with a compression factor of N/M, the sampling interval should be taken within the range of N/M and 2(N/M)-1 [10]. In Fig.…”

“…One of the random undersampling methods uses a continuous uniform distribution as a pseudorandom distribution with a certain interval sampling width. This approach can reduce the ADC power consumption [10]. However, an optimal distribution of the sampling series has not been proposed to further reduce the ADC power consumption, and it is thought that the distribution of the sampling series needs to be improved to achieve further reduction in ADC and wireless transmission power consumption.…”