Achieving Super-Resolution With Redundant Sensing

Objective: A new technique is presented to enhance the precision of the analog-to-digital (AD) and digital-to-analog (DA) conversion, which are fundamental operations of many biomedical information processing systems. In practice, the precision of these operations is always bounded, first by the random mismatch error occurred during system implementation, and subsequently by the intrinsic quantization error determined by the system architecture itself. Methods: Here, we derive a new mathematical interpretation of the previously proposed redundant sensing architecture that not only suppresses mismatch error but also allows achieving an effective resolution exceeding the system's intrinsic resolution, i.e., super-resolution (SR). SR is enabled by an endogenous property of redundant structures regarded as “code diffusion” where the references' value spreads into the neighbor sample space as a result of mismatch error. Results: Using Monte Carlo methods, we show a profound theoretical increase of 8-9 b effective resolution or 256-512× enhancement of precision on a 10-b device at 95% sample space. Conclusion: The proposed SR mechanism can be applied to substantially improve the precision of various AD and DA conversion processes beyond the system resource constraints. Significance: The concept opens the possibility for a wide range of applications in low-power fully integrated sensors and devices where the cost-accuracy tradeoff is inevitable. As a proof-of-concept demonstration, we point out an example where the proposed technique can be used to enhance the precision of an implantable neurostimulator design.