Hardware Reference
In-Depth Information
Fig. 5.14
The result of
device placement derived by
the baseline algorithm. The
output ports of reservoir
1
L
max
L
max
L
max
...
R
1
R
2
2 and the
heater are placed on the
boundary of the layout one by
one. R
1
and R
2
here represent
the output ports of two
reservoirs
7,DE1
...
Tabl e 5. 2
Comparison of the PCR biochips derived by proposed method and the baseline
algorithm
Proposed method
Baseline algorithm
Size of the
No.
Execution
Size of the
No.
Execution
Bioassay
Mixing ratio
biochip
electrodes
time (s)
biochip
electrodes
time (s)
1
2 W 3 W 5 W 7 W
11 11
44
35 C 36t
m
20 15
68
80 C 115t
m
11 W 13 W 87
(38.6)
(91.5)
[
27
]
2
51 W 51 W 41 W 4 W
13 11
49
80 C 79t
m
20 15
68
105 C 163t
m
4
W
5
W
351
W
5
(87.9)
(121.3)
[
34
]
3
7
W
14
W
11
11
6
30
25
C
12t
m
10
10
38
25
C
37t
m
[
35
]
(26.2)
(28.7)
*Numbers in parenthesis indicate execution time for t
m
D 0:1
We consider another mixing protocol called Bioassay 3 [
35
], which has three
input reagents/samples. The results are shown in Table
5.2
. In all the three cases,
the layout size, electrode count, and execution time of bioassays are significantly
improved compared to the baseline method.
Since these three laboratory bioassays have relatively low numbers of devices
that must be placed, the experimental results of the proposed algorithm are
optimized by Approach 1 introduced in Sect.
5.3.4
.
We further create three benchmarks that have a relatively large number of
reservoirs and apply Approach 1, Approach 2, and the baseline algorithm to
generate the resulting placements of the devices. Table
5.3
provides the results of
the simulation. It is important to notice that, when searching the optimal device-
placement results for Benchmarks 2 and 3, Approach 1 is impractical due to it high
computational complexity. In Approach 2, the parameters ˛ and ˇ in
5.8
are set as
1 and 10, respectively. From Table
5.3
, we find that Approach 2 can reduce the area
of the biochip by 64.1-72.9 % compared with the baseline algorithm.
