Abstract:
The bottleneck of recording multiple parallel channels with resource limited sensor circuits is the achievable data rate. Compressed Sensing (CS) is employed as a means o...Show MoreMetadata
Abstract:
The bottleneck of recording multiple parallel channels with resource limited sensor circuits is the achievable data rate. Compressed Sensing (CS) is employed as a means of compressing the recorded data with minimally added hardware overhead. When recording multiple channels using CS one can choose to implement CS for each channel separately or using a multi channel approach to compress all signals at once. This paper presents an empirical study of these two options in terms of achievable reconstruction performance depending on the number of channels and it also presents possible circuit implementations based on successive approximation register digital to analog converter (SAR ADC). The multi channel approach requires a certain minimum number of input channels to be feasible. Exceeding this minimum will increase the performance beyond the limits of the single channel implementation.
Date of Conference: 27-30 May 2018
Date Added to IEEE Xplore: 04 May 2018
ISBN Information:
Electronic ISSN: 2379-447X