Synthesis aware sample preparation techniques using random sample sets in DMFB

Abstract

Digital Microfluidic Biochips in recent years have emerged as a major feasible alternative platform for droplet based microfluidic operations within a miniaturized substrate. These devices find applications in the areas involving medical diagnostics, biochemical analysis and environmental toxicity monitoring issues. Formation of the sequencing graph for sample preparation followed by architecture level synthesis is the initial primary procedure for design automation of DMFBs. In this paper we proposed a technique for sample preparation using base solutions for a given random set of samples. The technique is developed to be synthesis aware in order to enable efficient use of resources in terms of modules for pipelined synthesis. The advantages of the proposed method is i) optimal resource utilization in terms of number of modules and samples and ii) efficient enhancement in overall execution time for the bioassay. The simulation is carried out with a given set of random samples as well as with a specified set of samples with Gaussian distribution. The results show encouraging improvements in i) number of sample utilizations, ii)number of mix-split operations iii)the formation of the number of waste droplets and iv) the actual number of modules used for execution of the overall bioassay.