Abstract:
Micro
uidics biochips are revolutionary devices in the eld of clinical diagnostics,
DNA analysis and molecular biology. Biochips involve various elds of science and
engineering i.e., physics, chemistry, biochemistry, nanotechnology, fabrication tech-
nology and computer science. It uses small volume of
uids on scale of micro to
nano liter for automatically carrying out the reactions needed for some biochemical
assays. In this report I have covered basic overview of biochips, existing algorithms
for sample preparation speci cally for dilution, mixing, and multiple droplet of sin-
gle target. Some application requires perticular sample repeatedy at the time of
asssy execution. So fast and e cient sample preparation process is require to gen-
erate multiple droplets of sample. Three approaches are presented in this report to
reduce the time to generate stream of droplets for bioassay having high demand of
sample at the time of assay execution. Simulation results show that new approach
is quite promising as compare to existing MMS and SRS algorithms to reduce the
time taken to prepare sample for such high demand.
First algorithm proposed, called KMS or K-Mixer Scheduling, utilizes K-Droplet
Mixer[3] to schedule mixing tree for multiple demand of single target generation.
This algorithm reduces total mix-split steps, by 74:6% than MMS and SRS but
compromises with storage requirement by 18:5% than SRS but still good by 25:1%
than MMS. To reduce the storage requirement modi ed KMS (m-KMS) schedules
mixing tree more then once with fraction f of total required demand D, fraction
f used to balance the total storage requirement U and total mixing operation Tms.
This new modi ed algorithm reduces total mix-split steps, by 79% and storage
requirement by 80:2%; 67:7% than MMS, SRS respectively. And nally we presented
KMS for mixing graph which used to schedule mixing graph. This algorithm reduces
total mix-split steps by 76% than MMS and SRS, storage requirement by 79:6%,
66:5% than MMS and SRS respectively.
