A defect tolerant high level synthesis method for digital microfluidic biochips based on the improved genetic algorithm

Digital microfluidic biochip (DMFB) is an attractive instrument for modern molecular biology and chemical measurements due to its flexible application and low fabrication cost. Due to the increasingly complex measurement is carried on the DMFB, such chips are more likely to face the risk of failure. So a fault tolerant design method for DMFB is needed. In this paper, we commit to solve the architecture level re-synthesis problem for DMFB when there are fault electrodes on the chip when the faults are detected and located. We proposed a defect tolerant high level synthesis for DMFB using the heuristic method of Improved Genetic Algorithm (IGA), which can solve the problem of resource binding, scheduling and placement at one time. Compared with the traditional randomly encode of the chromosome, we propose improved encode and decode method to solve the sequential constraint. Also, the method is robust when permanent faults happens. The simulation results showed our method is convergent, the IGA save 7% experiment time when there is no fault electrodes, 57% time while the opposite by comparing with the T-tree method and 3D defer decision method.