I
nte
rna
t
io
na
l J
o
urna
l o
f
Rec
o
nfig
ura
ble a
nd
E
m
bedd
e
d Sy
s
t
em
s
(
I
J
RE
S)
Vo
l.
10
,
No
.
2
,
J
u
l
y
2
0
2
1
,
p
p
.
7
7
~8
9
I
SS
N
: 2
0
8
9
-
4
8
6
4
,
DOI
: 1
0
.
1
1
5
9
1
/ijre
s
.
v
1
0
.
i2
.
p
p
7
7
-
89
77
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
r
es.ia
esco
r
e.
co
m
Co
st
-
ef
ficien
t
re
c
o
nfigura
ble geo
metrical bus
int
e
rc
o
nnection
sy
stem f
o
r ma
ny
-
co
re platforms
T
irum
a
le
Ra
m
esh
,
K
ha
lid
Abed
De
p
a
rtme
n
t
o
f
El
e
c
tri
c
a
l
&
Co
m
p
u
ter E
n
g
in
e
e
ri
n
g
a
n
d
C
o
m
p
u
ter S
c
ien
c
e
,
Ja
c
k
so
n
S
tate
Un
iv
e
rsity
,
USA
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Oct
2
0
,
2
0
2
0
R
ev
is
ed
Oct
2
9
,
2
0
2
0
Acc
ep
ted
Feb
1
5
,
2
0
2
1
S
y
ste
m
-
on
-
c
h
ip
(
S
o
C)
e
m
b
e
d
d
e
d
c
o
m
p
u
ti
n
g
p
latfo
rm
s
c
a
n
su
p
p
o
rt
a
wi
d
e
ra
n
g
e
o
f
n
e
x
t
g
e
n
e
ra
ti
o
n
e
m
b
e
d
d
e
d
a
rti
ficia
l
in
telli
g
e
n
c
e
a
n
d
o
t
h
e
r
c
o
m
p
u
tati
o
n
a
ll
y
i
n
ten
si
v
e
a
p
p
l
ica
ti
o
n
s
.
Th
e
se
p
latfo
rm
s
re
q
u
ire
c
o
st
e
ff
e
c
ti
v
e
in
terc
o
n
n
e
c
ti
o
n
n
e
tw
o
r
k
.
N
e
two
rk
-
on
-
c
h
ip
h
a
s
b
e
e
n
w
id
e
ly
u
se
d
to
d
a
y
fo
r
o
n
-
c
h
ip
in
terc
o
n
n
e
c
ti
o
n
.
Ho
we
v
e
r,
it
is
st
il
l
c
o
n
sid
e
re
d
e
x
p
e
n
si
v
e
fo
r
larg
e
sy
ste
m
siz
e
s.
As
f
u
ll
b
u
s
-
b
a
se
d
in
terc
o
n
n
e
c
ti
o
n
h
a
s
h
i
g
h
n
u
m
b
e
r
o
f
b
u
s
c
o
n
n
e
c
ti
o
n
s
,
re
d
u
c
e
d
b
u
s
c
o
n
n
e
c
ti
o
n
s
m
ig
h
t
o
ffe
r
c
o
n
sid
e
ra
b
le
imp
lem
e
n
tatio
n
e
c
o
n
o
m
ies
wit
h
re
lativ
e
l
y
sm
a
ll
d
e
sig
n
c
o
st
fo
r
field
p
ro
g
ra
m
m
a
b
le
g
a
te
a
rra
y
s
(
F
P
G
A
s
)
b
a
se
d
e
m
b
e
d
d
e
d
p
latf
o
rm
s.
I
n
t
h
is p
a
p
e
r,
we
p
ro
p
o
se
a
c
o
st
e
fficie
n
t
g
e
n
e
ra
li
z
e
d
re
c
o
n
f
ig
u
ra
b
le
b
u
s
-
b
a
se
d
in
terc
o
n
n
e
c
ti
o
n
f
o
r
m
a
n
y
-
c
o
re
sy
ste
m
wit
h
re
d
u
c
e
d
n
u
m
b
e
r
o
f
b
u
s
c
o
n
n
e
c
ti
o
n
s
.
We
g
e
n
e
ra
li
z
e
t
h
e
s
y
ste
m
with
b
n
u
m
b
e
r
o
f
i
n
terc
o
n
n
e
c
t
b
u
se
s
in
w
h
ich
=
{
,
}
/
wh
e
re
n
is
th
e
n
u
m
b
e
r
o
f
p
r
o
c
e
ss
o
r
c
o
re
s,
m
is
th
e
n
u
m
b
e
r
o
f
m
e
m
o
ry
-
m
o
d
u
les
a
n
d
k
is
th
e
g
e
n
e
ra
l
b
u
s red
u
c
ti
o
n
f
a
c
to
r.
We
p
re
se
n
t
fo
u
r
g
e
o
m
e
tri
c
a
l
in
terc
o
n
n
e
c
t
c
o
n
fig
u
ra
ti
o
n
s
a
n
d
p
ro
v
id
e
t
he
ir
c
h
a
ra
c
teriz
a
ti
o
n
in
term
s
o
f
m
e
m
o
ry
b
a
n
d
wid
t
h
,
c
o
st
p
e
r
b
a
n
d
wi
d
th
a
n
d
b
u
s
fa
u
lt
to
lera
n
c
e
fo
r
v
a
ri
o
u
s
sy
s
tem
siz
e
s.
Ou
r
re
su
lt
s
sh
o
w
th
a
t
th
e
se
c
o
n
fig
u
ra
ti
o
n
s
p
ro
v
id
e
re
d
u
c
e
d
c
o
st
p
e
r
b
a
n
d
wid
t
h
a
n
d
c
a
n
a
c
h
ie
v
e
h
i
g
h
e
r
sy
ste
m
th
ro
u
g
h
p
u
t
wit
h
b
u
s
c
a
c
h
e
.
K
ey
w
o
r
d
s
:
Ar
tific
ial
in
tellig
en
ce
C
o
s
t p
er
b
an
d
wid
th
E
d
g
e
co
m
p
u
tin
g
Geo
m
etr
ical
b
u
s
co
n
n
ec
tio
n
s
Ma
n
y
-
co
r
e
em
b
ed
d
ed
On
-
ch
ip
in
ter
c
o
n
n
ec
tio
n
R
ec
o
n
f
ig
u
r
ab
le
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Kh
alid
Ab
ed
Dep
ar
tm
en
t o
f
E
lectr
ical
an
d
C
o
m
p
u
ter
E
n
g
in
ee
r
in
g
an
d
C
o
m
p
u
ter
Scien
ce
J
ac
k
s
o
n
State
Un
iv
er
s
ity
1
4
0
0
J
o
h
n
R
L
y
n
ch
Stre
et
(
J
SU Bo
x
1
7
0
9
8
)
,
J
ac
k
s
o
n
,
MS.
3
9
2
1
7
,
Un
ited
States
E
m
ail:
k
h
alid
.
h
.
a
b
ed
@
js
u
m
s
.
ed
u
1.
I
NT
RO
D
UCT
I
O
N
Mu
lti
-
co
r
e
p
r
o
ce
s
s
o
r
d
esig
n
s
ar
e
n
o
w
s
ee
in
g
b
ig
s
h
if
t
[
1
]
to
war
d
s
m
an
y
-
c
o
r
e
co
m
p
u
tin
g
[
2
]
.
Ma
n
y
-
co
r
e
p
r
o
ce
s
s
o
r
s
y
s
tem
s
ar
e
n
o
w
tr
en
d
in
g
as
a
p
latf
o
r
m
f
o
r
m
ass
iv
e
p
ar
allel
co
m
p
u
tin
g
[
3
]
an
d
ar
e
also
h
a
d
b
ee
n
u
s
ed
as
a
co
-
p
r
o
ce
s
s
o
r
f
o
r
a
m
u
lti
-
co
r
e
s
y
s
tem
[
4
]
.
I
n
r
ec
en
t
y
ea
r
s
,
we
s
ee
an
em
e
r
g
en
ce
o
f
ex
tr
em
e
co
m
p
u
tin
g
[
5
]
f
o
r
b
ig
d
ata.
So
m
e
ar
ch
i
tectu
r
al
m
o
d
els
f
o
r
m
an
y
-
co
r
e
p
r
o
ce
s
s
o
r
s
y
s
tem
s
[
6
,
7
]
an
d
p
er
f
o
r
m
an
ce
im
p
r
o
v
em
en
t
te
ch
n
iq
u
es
[
8
]
f
o
r
m
u
lti
-
co
r
e
h
av
e
b
ee
n
d
ev
el
o
p
ed
.
Ma
n
y
-
co
r
e
co
m
p
u
ti
n
g
is
g
ain
in
g
in
ter
ests
f
o
r
ar
tific
ial
in
tellig
en
ce
(
AI
)
b
ased
d
e
f
en
s
e
ap
p
licatio
n
s
[
9
]
.
E
d
g
e
co
m
p
u
tin
g
[
10
]
is
n
o
w
co
n
s
id
er
ed
as
a
r
elativ
ely
a
n
ew
p
ar
ad
i
g
m
wh
e
r
e
th
e
co
m
p
u
tatio
n
al
r
eso
u
r
ce
s
ar
e
p
lace
d
at
th
e
ed
g
e
o
f
th
e
n
etwo
r
k
.
Use
o
f
m
u
lti
-
co
r
e
i
s
g
ain
in
g
in
ter
ests
f
o
r
ed
g
e
co
m
p
u
tin
g
[
1
1
]
as
d
ata
tr
an
s
f
er
b
etwe
en
co
r
es
is
p
o
wer
d
em
a
n
d
in
g
an
d
r
eq
u
ir
e
a
v
er
y
co
m
p
le
x
co
n
n
ec
tio
n
in
f
r
astru
ctu
r
e.
W
ith
i
n
cr
ea
s
ed
v
er
y
l
ar
g
e
-
s
ca
le
i
n
teg
r
atio
n
(
VL
SI)
d
en
s
ity
,
h
e
ter
o
g
en
eo
u
s
m
an
y
-
c
o
r
e
b
ased
ch
ip
m
u
ltip
r
o
ce
s
s
o
r
s
(
C
MP)
[
12
]
f
o
r
b
ig
d
ata
ar
e
o
n
th
e
r
is
e.
Fo
r
t
h
ese
ap
p
li
ca
tio
n
s
r
u
n
n
in
g
o
n
m
a
n
y
-
c
o
r
e
p
latf
o
r
m
s
,
we
r
eq
u
ir
e
c
o
s
t
ef
f
icien
t
o
n
-
c
h
ip
in
ter
co
n
n
ec
tio
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
9
-
4
8
6
4
I
n
t J Reco
n
f
ig
u
r
a
b
le
&
E
m
b
ed
d
ed
Sy
s
t,
Vo
l.
10
,
No
.
2
,
J
u
ly
2
0
2
1
:
7
7
–
89
78
Mo
s
t
o
f
th
e
wo
r
k
to
d
ay
f
o
r
o
n
-
ch
i
p
in
ter
co
n
n
ec
tio
n
is
n
etwo
r
k
-
on
-
c
h
i
p
(
No
C
)
[
1
3
-
1
5
]
.
T
h
ey
g
en
er
ally
o
f
f
e
r
m
o
r
e
d
eg
r
ee
o
f
f
r
ee
d
o
m
th
an
b
u
s
-
b
ased
s
y
s
tem
s
.
No
C
h
as
h
i
g
h
o
v
er
h
ea
d
d
u
e
to
p
ac
k
etiza
tio
n
an
d
m
u
lti
-
h
o
p
r
o
u
tin
g
,
wh
ich
af
f
ec
ts
p
h
y
s
ical
ar
ea
,
p
o
wer
an
d
laten
cy
lead
in
g
to
in
cr
ea
s
e
d
co
s
t
an
d
cr
ea
tes
ch
allen
g
es
f
o
r
ac
h
ie
v
in
g
h
i
g
h
p
er
f
o
r
m
an
ce
.
C
u
r
r
e
n
t
No
C
r
esear
ch
es
co
n
tin
u
e
to
ad
d
r
ess
th
ese
ch
allen
g
es.
I
n
r
ec
en
t
y
ea
r
s
,
s
o
m
e
n
o
v
el
No
C
ar
ch
it
ec
tu
r
es
[
1
6
-
2
1
]
wer
e
p
r
o
p
o
s
ed
.
I
n
[
1
6
]
,
b
u
f
f
er
less
cir
cu
it
s
witch
ed
No
C
was
ad
d
r
ess
ed
f
o
r
b
o
th
co
s
t
a
n
d
e
n
er
g
y
ef
f
icien
cy
.
A
h
y
b
r
i
d
No
C
was
p
r
o
p
o
s
ed
in
[
1
7
]
.
R
eliab
ilit
y
o
f
No
C
r
o
u
ter
was
f
o
cu
s
ed
in
[
1
8
]
.
A
f
ield
p
r
o
g
r
am
m
a
b
le
g
at
e
ar
r
ay
(
FP
GA)
b
ased
s
o
lu
tio
n
tar
g
etin
g
a
r
ea
o
p
tim
izatio
n
f
o
r
im
p
r
o
v
e
d
n
etwo
r
k
p
er
f
o
r
m
a
n
ce
u
s
in
g
s
ev
er
al
tech
n
iq
u
es
was
p
r
o
p
o
s
ed
r
ec
en
tly
i
n
[
1
9
]
.
T
h
er
e
was
s
o
m
e
r
ec
e
n
t
wo
r
k
o
n
r
o
u
ter
less
No
C
[
20
]
u
s
in
g
s
m
ar
t
on
-
ch
i
p
wir
in
g
r
eso
u
r
c
e
m
an
ag
em
en
t
f
o
r
r
ed
u
cin
g
co
s
t.
I
n
[
2
1
]
,
an
e
f
f
icien
t
cr
o
s
s
b
ar
s
witch
im
p
lem
en
tatio
n
f
o
r
No
C
was
p
r
o
p
o
s
ed
.
Hete
r
o
g
en
e
o
u
s
co
m
p
u
tin
g
[
2
2
]
ex
is
ts
to
d
a
y
i
n
th
e
f
o
r
m
o
f
f
u
n
ctio
n
ally
d
iv
er
s
e
co
m
p
u
tatio
n
al
d
ev
ices,
m
em
o
r
y
s
y
s
te
m
s
an
d
in
clu
d
es
h
eter
o
g
e
n
eity
in
in
ter
co
n
n
ec
tio
n
.
Ma
ch
in
e
lear
n
in
g
is
g
ain
in
g
i
n
ter
ests
in
p
r
o
ce
s
s
o
r
af
f
in
ity
ch
ar
ac
ter
izatio
n
f
o
r
f
u
n
ctio
n
a
lly
d
iv
er
s
e
h
eter
o
g
en
o
u
s
d
ev
ices
[
2
3
]
.
T
h
e
r
e
is
s
o
m
e
r
ec
en
t
wo
r
k
[
2
4
]
o
n
o
p
tim
izatio
n
o
f
m
u
lti
-
co
r
e
s
y
s
tem
s
u
s
in
g
m
a
c
h
in
e
lea
r
n
in
g
.
Sm
aller
s
ize
m
an
y
-
c
o
r
e
p
r
o
ce
s
s
o
r
s
with
No
C
o
n
FP
GA
s
ar
e
b
u
ilt
[
2
5
]
.
T
h
e
r
e
is
s
o
m
e
r
ec
en
t h
y
b
r
id
No
C
wo
r
k
[
2
6
]
th
at
u
tili
ze
d
b
u
s
-
b
ased
i
n
ter
co
n
n
ec
tio
n
.
A
ch
iev
in
g
e
n
er
g
y
ef
f
icien
t
o
n
-
ch
ip
in
ter
c
o
n
n
ec
tio
n
at
r
ed
u
c
ed
ar
ea
c
o
s
t
f
o
r
th
e
s
ca
le
o
f
h
u
n
d
r
e
d
s
o
f
m
an
y
-
co
r
es
is
a
m
aj
o
r
c
h
allen
g
e
.
B
us
-
b
ased
in
ter
co
n
n
ec
tio
n
g
e
n
er
ally
c
o
n
s
id
er
ed
s
im
p
le
r
in
d
esig
n
,
en
er
g
y
ef
f
icien
t
an
d
f
a
u
lt
to
ler
an
t
th
a
n
No
C
;
b
u
t
lack
s
s
ca
lab
ilit
y
f
o
r
l
ar
g
e
s
y
s
tem
s
ize
.
So
m
e
V
L
SI
ex
p
er
ts
i
n
th
e
p
ast
ar
g
u
ed
th
at
m
etal/wir
in
g
ar
e
ch
ea
p
an
d
p
len
tifu
l
an
d
we
ar
e
n
o
lo
n
g
er
p
in
-
lim
ited
th
u
s
f
av
o
r
in
g
b
u
s
-
b
ased
o
n
-
ch
ip
in
ter
c
o
n
n
ec
tio
n
[
2
7
]
.
Fo
r
m
u
lti
-
co
r
e
p
r
o
ce
s
s
o
r
s
y
s
tem
s
,
p
e
r
f
o
r
m
an
ce
m
ay
n
o
t
b
e
m
ax
im
ized
b
y
e
v
en
u
s
in
g
h
ig
h
est
b
a
n
d
w
id
th
o
n
-
ch
ip
in
ter
co
n
n
ec
ts
to
p
o
lo
g
y
as
it
c
o
n
s
u
m
es
p
o
wer
an
d
ar
ea
r
eso
u
r
ce
s
[
2
8
]
.
As
f
u
r
th
er
a
r
g
u
e
d
in
[
2
8
]
,
in
c
r
ea
s
in
g
th
e
n
u
m
b
er
o
f
m
u
lti
-
co
r
es
p
lace
s
h
ig
h
er
in
te
r
co
n
n
ec
t
b
an
d
wid
th
d
em
an
d
wh
ich
d
ec
r
ea
s
es
th
e
av
ailab
le
s
ilico
n
r
ea
l
e
s
tate
f
o
r
lar
g
e
s
ize
m
u
lti
-
co
r
es
.
T
h
u
s
,
we
b
eliev
e
th
at
b
y
tak
in
g
ad
v
a
n
tag
e
o
f
cu
r
r
en
t
VL
SI
d
en
s
ity
,
s
ilico
n
r
ea
l
estate
ca
n
b
e
s
till
b
e
m
an
ag
e
d
f
o
r
s
m
all
s
ize
m
an
y
-
co
r
es
co
m
b
in
ed
with
r
ed
u
ce
d
co
s
t
in
ter
co
n
n
ec
tio
n
.
T
h
is
m
o
tiv
ated
u
s
to
lo
o
k
in
to
s
im
p
ler
r
e
d
u
ce
d
c
o
s
t
bus
-
b
ased
on
-
c
h
ip
in
ter
c
o
n
n
ec
tio
n
s
o
lu
tio
n
f
o
r
m
an
y
-
co
r
es
th
at
ca
n
ac
h
iev
e
s
am
e
b
a
n
d
wid
th
an
d
o
f
f
er
g
o
o
d
b
u
s
f
au
lt to
ler
an
ce
f
o
r
m
o
d
er
ate
n
u
m
b
er
o
f
co
r
es (
16
-
12
8
)
.
O
u
r
m
ain
co
n
tr
ib
u
tio
n
s
o
f
th
e
p
a
p
e
r
ar
e
as f
o
llo
ws:
−
Pro
p
o
s
e
a
g
en
e
r
alize
d
r
ec
o
n
f
ig
u
r
ab
le
m
a
n
y
-
c
o
r
e
s
y
s
tem
with
co
s
t
ef
f
ec
tiv
e
g
eo
m
etr
ic
al
b
u
s
on
-
c
h
i
p
in
ter
co
n
n
ec
tio
n
co
n
f
ig
u
r
ati
o
n
s
ex
ten
d
ed
f
r
o
m
o
u
r
ea
r
lier
wo
r
k
[
2
9
]
.
−
P
r
esen
t
th
e
b
u
s
ar
b
itra
tio
n
al
g
o
r
ith
m
f
o
r
t
h
ese
co
n
f
i
g
u
r
ati
o
n
s
an
d
p
r
o
v
id
e
a
c
o
m
p
r
eh
e
n
s
iv
e
s
y
s
tem
ch
ar
ac
ter
izatio
n
in
ter
m
s
o
f
m
em
o
r
y
b
a
n
d
wid
th
,
c
o
s
t
p
er
b
an
d
wid
th
,
b
u
s
f
au
lt
to
ler
an
ce
an
d
s
y
s
tem
th
r
o
u
g
h
p
u
t w
ith
b
u
s
ca
ch
e
a
d
d
ed
o
n
ea
c
h
b
u
s
lin
e
.
−
Pre
s
en
t
d
etailed
r
esu
lts
an
d
d
i
s
cu
s
s
th
ese
r
esu
lts
.
−
E
s
tim
ate
co
s
t o
f
th
ese
co
n
f
ig
u
r
atio
n
s
in
co
m
p
ar
is
o
n
to
a
n
ex
am
p
le
cir
cu
it swit
ch
ed
r
o
u
ter
.
−
P
r
o
v
id
e
co
n
clu
s
io
n
an
d
p
r
esen
t so
m
e
in
s
ig
h
t in
to
f
u
tu
r
e
r
ese
ar
ch
.
2.
SYST
E
M
ARCH
I
T
E
CT
U
R
E
AND
CO
NIGU
RAT
I
O
NS
2
.
1
.
G
ener
a
lized
re
co
nfig
ura
ble
m
a
ny
-
co
re
s
y
s
t
em
pla
t
f
o
rm
Fro
m
th
e
ea
r
lier
wo
r
k
o
n
m
u
ltip
le
b
u
s
s
y
s
tem
[
3
0
]
,
it
h
as
b
ee
n
o
b
s
er
v
ed
th
at
b
y
u
s
in
g
n
u
m
b
er
o
f
b
u
s
es
eq
u
al
to
o
n
e
-
h
al
f
o
f
t
h
e
n
u
m
b
e
r
o
f
m
em
o
r
y
-
m
o
d
u
les
o
r
p
r
o
ce
s
s
o
r
s
,
we
ca
n
ac
h
iev
e
a
m
em
o
r
y
b
an
d
wid
th
with
in
2
5
%
o
f
th
e
cr
o
s
s
b
ar
b
an
d
wid
t
h
.
Fo
r
co
m
p
lete
b
u
s
co
n
n
ec
tio
n
s
,
all
m
em
o
r
y
-
m
o
d
u
les
co
n
n
e
cted
to
all
b
u
s
es
.
In
c
r
ea
s
e
in
b
u
s
es
in
cu
r
s
h
ig
h
n
u
m
b
er
o
f
b
u
s
c
o
n
n
ec
tio
n
s
an
d
co
s
t.
E
ar
lies
t
wo
r
k
o
n
r
ed
u
ce
d
b
u
s
co
n
n
ec
tio
n
s
ch
em
es
[
3
1
]
s
h
o
wed
g
en
e
r
al
th
eo
r
em
s
an
d
p
r
o
p
er
ties
f
o
r
th
e
s
ch
em
e.
W
ith
r
ed
u
ctio
n
in
th
e
n
u
m
b
er
o
f
b
u
s
es
an
d
b
u
s
co
n
n
ec
tio
n
s
,
w
e
p
r
o
p
o
s
e
a
r
ec
o
n
f
ig
u
r
a
b
le
m
an
y
-
co
r
e
p
latf
o
r
m
with
n
co
r
es
,
m
m
em
o
r
y
-
m
o
d
u
les,
b
b
u
s
es
an
d
k
b
u
s
r
ed
u
ctio
n
f
a
cto
r
.
Fig
u
r
e
1
s
h
o
ws
th
e
s
y
s
tem
ar
ch
itectu
r
e
th
a
t
in
clu
d
es b
u
s
ca
ch
e
at
ea
c
h
b
u
s
lin
e
an
d
r
ec
o
n
f
i
g
u
r
ab
le
c
o
n
tr
o
l.
=
m
i
n
{
,
}
(
1
)
Fig
u
r
e
1
.
R
ec
o
n
f
ig
u
r
a
b
le
m
an
y
-
co
r
e
s
y
s
tem
p
latf
o
r
m
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Reco
n
f
ig
u
r
a
b
le
&
E
m
b
ed
d
ed
Sy
s
t
I
SS
N:
2089
-
4
8
6
4
C
o
s
t
-
efficien
t reco
n
fig
u
r
a
b
le
g
eo
metrica
l b
u
s
in
terco
n
n
ec
ti
o
n
s
ystem
…
(
Tir
u
ma
le
R
a
mes
h
)
79
2
.
2
.
G
eo
m
e
t
rica
l bus
inte
rc
o
nn
ec
t
io
n c
o
nfig
ura
t
io
ns
Ou
r
ea
r
lier
wo
r
k
in
[
2
9
]
,
we
c
o
m
p
lim
en
ted
[
3
1
]
an
d
p
r
esen
t
ed
a
g
en
e
r
alize
d
s
y
s
tem
ar
ch
it
ec
tu
r
e
an
d
ch
ar
ac
ter
izatio
n
.
In
th
is
p
ap
er
we
s
u
p
p
lem
en
t
ex
p
an
d
s
o
u
r
ea
r
lier
wo
r
k
[
29
]
an
d
p
r
o
p
o
s
e
f
o
u
r
d
is
tin
ct
g
eo
m
etr
ical
b
u
s
co
n
f
ig
u
r
atio
n
s
:
i)
.
Gr
o
u
p
R
h
o
m
b
ic
2
(
GR
2
)
,
ii).
Gr
o
u
p
R
h
o
m
b
ic
4
(
GR
4
)
,
iii).
Hier
ar
ch
ical
R
h
o
m
b
ic
(
HR
)
an
d
iv
)
.
Qu
a
d
r
an
t
R
h
o
m
b
ic
(
QR
)
.
W
e
p
r
o
v
id
e
a
co
m
p
r
e
h
en
s
iv
e
s
y
s
tem
ch
ar
ac
ter
izatio
n
f
o
r
t
h
ese
co
n
f
ig
u
r
atio
n
s
.
Alth
o
u
g
h
in
g
en
e
r
ality
,
we
s
tated
a
b
u
s
r
e
d
u
ctio
n
f
ac
to
r
in
(
1
)
,
we
u
s
ed
=
2
th
r
o
u
g
h
o
u
t
t
h
is
p
ap
er
.
We
co
n
s
id
er
ed
r
h
o
m
b
ic
as
a
g
eo
m
etr
ical
p
atter
n
b
ase
to
d
ef
in
e
th
ese
f
o
u
r
co
n
f
ig
u
r
atio
n
s
as
r
h
o
m
b
ic
was
co
n
s
id
er
ed
m
o
s
t
co
s
t
-
ef
f
ec
ti
v
e
to
p
o
lo
g
y
[
2
9
]
.
Ho
we
v
er
,
in
g
en
er
al,
an
y
o
f
th
e
o
th
er
g
e
o
m
etr
ical
p
atter
n
p
r
esen
ted
in
[
2
9
]
co
u
ld
b
e
u
s
ed
a
s
a
b
ase
.
I
n
Fig
u
r
e
2
,
Fig
u
r
e
3
an
d
Fig
u
r
e
4
,
all
co
n
n
ec
tio
n
s
m
a
r
k
ed
as “
x
”
r
ef
er
s
to
th
e
b
u
s
es n
u
m
b
er
ed
o
n
i
ts
lef
t.
−
GR
2
:
Me
m
o
r
y
-
m
o
d
u
les
an
d
b
u
s
es
d
iv
id
ed
in
t
o
two
g
r
o
u
p
s
co
n
n
ec
ted
in
r
h
o
m
b
ic
p
at
ter
n
.
Fig
u
r
e
2
s
h
o
ws th
e
co
n
f
ig
u
r
atio
n
.
All p
r
o
ce
s
s
o
r
c
o
r
es c
o
n
n
ec
ted
to
all
b
u
s
es.
−
GR
4
:
Me
m
o
r
y
-
m
o
d
u
les
an
d
b
u
s
es
d
iv
id
ed
i
n
to
f
o
u
r
g
r
o
u
p
s
co
n
n
ec
ted
in
r
h
o
m
b
ic
p
att
er
n
.
Fig
u
r
e
3
s
h
o
ws th
e
co
n
f
ig
u
r
atio
n
.
All p
r
o
ce
s
s
o
r
c
o
r
es c
o
n
n
ec
ted
to
all
b
u
s
es.
−
HR
:
Me
m
o
r
y
-
m
o
d
u
les
an
d
c
o
r
es
co
n
n
ec
ted
in
h
ie
r
ar
ch
ical
b
u
s
s
y
s
tem
[
3
2
]
with
m
o
d
if
ic
atio
n
s
.
C
o
r
es
ar
e
co
n
n
ec
ted
in
two
g
r
o
u
p
s
in
lev
el
1
an
d
m
em
o
r
y
-
m
o
d
u
l
es
co
n
n
ec
ted
in
r
h
o
m
b
ic
p
att
er
n
in
lev
el
2
.
Pro
ce
s
s
o
r
c
o
r
e
an
d
m
em
o
r
y
b
u
s
es
ar
e
in
ter
co
n
n
ec
ted
.
F
ig
u
r
e
4
s
h
o
ws
b
o
th
m
e
m
o
r
y
-
m
o
d
u
le
an
d
p
r
o
ce
s
s
o
r
co
r
e
co
n
n
ec
tio
n
s
.
−
QR
:
Me
m
o
r
y
-
m
o
d
u
les
an
d
b
u
s
es
d
iv
id
ed
in
to
f
o
u
r
q
u
ad
r
an
t
s
with
ea
ch
q
u
ad
r
a
n
t
co
n
n
ec
te
d
i
n
r
h
o
m
b
ic
p
atter
n
.
Fig
u
r
e
5
s
h
o
ws th
e
m
em
o
r
y
-
m
o
d
u
le
co
n
n
ec
tio
n
s
Gr
o
u
p
an
d
q
u
a
d
r
an
t
r
h
o
m
b
ic
co
n
f
ig
u
r
atio
n
s
ar
e
tig
h
tly
c
o
u
p
led
as
b
o
t
h
co
r
e
a
n
d
m
e
m
o
r
y
b
u
s
co
n
n
ec
tio
n
s
ar
e
o
n
th
e
s
am
e
s
et
o
f
b
u
s
es.
Hier
ar
ch
ical
r
h
o
m
b
ic
is
a
lo
o
s
ely
co
u
p
led
co
n
f
i
g
u
r
atio
n
s
in
ce
th
ey
h
av
e
s
ep
ar
ate
co
r
e
an
d
m
em
o
r
y
b
u
s
co
n
n
ec
tio
n
s
.
I
t
was
s
h
o
wn
in
[
3
1
]
th
at
f
o
r
an
y
r
h
o
m
b
ic
co
n
n
ec
tio
n
s
,
th
e
lo
wer
b
o
u
n
d
f
o
r
n
u
m
b
er
o
f
m
em
o
r
y
-
m
o
d
u
les
co
n
n
ec
ted
t
o
ea
ch
b
u
s
is
eq
u
al
to
(
−
+
1
)
to
ass
ig
n
b
u
s
es
to
m
em
o
r
y
-
m
o
d
u
les.
I
n
Fig
u
r
es
2
th
r
o
u
g
h
5
b
elo
w
s
h
o
ws
th
ese
co
n
f
ig
u
r
atio
n
s
.
M1
-
M1
6
is
m
em
o
r
y
-
m
o
d
u
les
an
d
P1
-
P8
is
p
r
o
ce
s
s
o
r
co
r
es.
W
e
ex
p
licitly
s
h
o
wn
p
r
o
ce
s
s
o
r
c
o
r
e
c
o
n
n
ec
tio
n
s
o
n
ly
f
o
r
h
ier
ar
ch
ical
co
n
f
ig
u
r
atio
n
.
Fig
u
r
e
2
.
Gr
o
u
p
r
h
o
m
b
ic
2
(
G
R
2
)
f
o
r
n
=m
=8
,
b
=4
Fig
u
r
e
3
.
Gr
o
u
p
r
h
o
m
b
ic
4
(
G
R
4
)
f
o
r
n
=m
=1
6
,
b
=8
Fig
u
r
e
4
.
Hier
ar
c
h
ical
r
h
o
m
b
i
c
(
HR
)
f
o
r
n
=m
=8
,
b
=4
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
9
-
4
8
6
4
I
n
t J Reco
n
f
ig
u
r
a
b
le
&
E
m
b
ed
d
ed
Sy
s
t,
Vo
l.
10
,
No
.
2
,
J
u
ly
2
0
2
1
:
7
7
–
89
80
Fig
u
r
e
5
.
Qu
a
d
r
an
t
r
h
o
m
b
ic
(
QR
)
f
o
r
n
=m
=8
,
b
=4
T
ab
le
1
s
u
m
m
ar
izes
th
e
to
t
al
co
s
t
o
f
in
ter
co
n
n
ec
tio
n
s
.
T
h
e
f
ir
s
t
ter
m
is
th
e
p
r
o
c
ess
o
r
co
r
e
co
n
n
ec
tio
n
co
s
t,
th
e
s
ec
o
n
d
te
r
m
is
th
e
m
e
m
o
r
y
co
n
n
ec
tio
n
co
s
t
an
d
th
e
th
ir
d
ter
m
is
th
e
c
o
r
e
to
m
em
o
r
y
b
u
s
in
ter
-
lev
el
co
n
n
ec
tio
n
c
o
s
t
f
o
r
h
ier
ar
ch
ical
r
h
o
m
b
ic.
As
o
b
s
er
v
ed
f
r
o
m
T
a
b
le
1
,
g
r
o
u
p
r
h
o
m
b
ic
h
as
r
ed
u
ce
d
n
u
m
b
er
o
f
b
u
s
c
o
n
n
ec
tio
n
s
co
m
p
ar
ed
to
r
eg
u
la
r
r
h
o
m
b
ic
[
2
9
]
.
Q
u
ad
r
an
t
r
h
o
m
b
ic
h
as
s
lig
h
tly
h
i
g
h
er
n
u
m
b
er
o
f
b
u
s
co
n
n
ec
tio
n
s
th
an
r
h
o
m
b
ic
b
u
t
is
in
h
er
en
tly
f
au
lt
to
ler
an
t
to
cr
itical
b
u
s
f
au
lts
(
ex
p
lain
ed
in
s
ec
tio
n
3
.
6
)
.
All
f
o
u
r
c
o
n
f
ig
u
r
atio
n
s
h
av
e
r
e
d
u
ce
d
n
u
m
b
er
o
f
b
u
s
co
n
n
ec
tio
n
s
co
m
p
a
r
ed
to
co
m
p
lete
b
u
s
co
n
n
ec
tio
n
s
.
Fig
u
r
e
6
s
h
o
ws
t
h
e
av
er
ag
e
c
o
s
t
s
av
in
g
s
.
T
h
e
co
s
t
s
av
in
g
s
r
an
g
es
f
r
o
m
2
4
%
to
4
2
%
ac
r
o
s
s
all
s
y
s
tem
s
izes.
T
ab
le
1
.
C
o
s
t o
f
in
ter
c
o
n
n
ec
ti
o
n
s
(
n
o
o
f
c
o
n
n
ec
tio
n
s
)
C
o
n
n
e
c
t
i
o
n
T
o
p
o
l
o
g
i
e
s
T
o
t
a
l
C
o
st
S
y
st
e
m Si
z
e
16
32
64
1
2
8
C
o
m
p
l
e
t
e
b
u
s
c
o
n
n
e
c
t
i
o
n
s
[
+
]
2
5
6
1
0
2
4
4
0
9
6
1
6
3
8
4
R
e
g
u
l
a
r
r
h
o
mb
i
c
[
2
9
]
.
+
[
−
+
1
]
2
0
0
7
8
4
3
1
0
4
1
2
3
5
2
G
r
o
u
p
R
h
o
m
b
i
c
2
.
+
[
2
−
(
2
−
1
)
]
1
6
8
6
5
6
2
5
9
2
1
0
3
0
4
G
r
o
u
p
R
h
o
m
b
i
c
4
.
+
[
4
−
(
4
−
1
)
]
1
5
2
5
9
2
2
3
3
6
9
2
8
0
H
i
e
r
a
r
c
h
i
c
a
l
R
h
o
m
b
i
c
.
2
+
[
−
+
1
]
+
.
2
0
0
7
8
4
3
1
0
4
1
2
3
5
2
H
i
e
r
a
r
c
h
i
c
a
l
R
h
o
m
b
i
c
.
+
2
[
2
−
(
2
−
1
)
]
2
0
8
8
0
0
3
1
3
6
1
2
4
1
6
Fig
u
r
e
6
.
Av
e
r
ag
e
co
s
t sav
in
g
s
2
.
3
.
Sy
s
t
e
m
re
co
nfig
ura
t
io
n
R
ec
o
n
f
ig
u
r
atio
n
C
o
n
tr
o
l
R
eg
is
ter
(
R
C
R
)
s
h
o
wn
in
Fig
u
r
e
7
f
ac
ilit
ates
r
ec
o
n
f
ig
u
r
atio
n
o
f
th
e
s
y
s
tem
f
o
r
p
lu
r
ality
o
f
p
r
o
ce
s
s
o
r
co
r
e
s
,
m
em
o
r
y
-
m
o
d
u
les
an
d
b
u
s
es
co
n
n
ec
ted
to
th
e
s
y
s
tem
an
d
h
av
e
th
e
f
o
llo
win
g
f
u
n
ctio
n
s
:
−
Switch
es SP,
S
M
an
d
SB
:
C
o
n
n
ec
ts
co
r
es,
m
em
o
r
y
-
m
o
d
u
les
an
d
b
u
s
es to
th
e
s
y
s
tem
r
esp
ec
tiv
ely
.
−
R
eg
is
ter
s
S
PR
,
SM
R
,
an
d
SB
R
:
C
o
n
tr
o
ls
SM,
SP
,
an
d
SB
r
esp
ec
tiv
ely
.
Fo
r
e
x
am
p
le,
if
SP
(
i
)
=1
,
th
en
a
co
r
e
is
co
n
n
ec
ted
to
th
e
s
y
s
tem
.
Similar
ly
,
SM
(
j
)
=
1
co
n
n
ec
ts
a
m
e
m
o
r
y
j
an
d
SB
(
k
)
=
1
co
n
n
ec
ts
a
b
u
s
k
to
th
e
s
y
s
tem
.
−
R
eg
is
ter
s
S
C
P
R
an
d
SC
MR:
C
o
n
tr
o
ls
th
e
r
ec
o
n
f
ig
u
r
atio
n
o
f
th
e
i
n
ter
co
n
n
ec
tio
n
t
o
co
n
n
ec
t
a
co
r
e
o
r
m
em
o
r
y
t
o
a
s
p
ec
if
ic
b
u
s
.
Fo
r
ex
am
p
le,
m
em
o
r
y
“
1
”
co
n
n
ec
t
ed
to
b
u
s
“1
”
f
o
r
SCM
(
1
,
1
)
=
1
.
−
On
a
s
in
g
le
b
u
s
f
au
lt,
th
e
in
ter
co
n
n
ec
tio
n
is
r
ec
o
n
f
i
g
u
r
e
d
an
d
u
p
d
ates
th
e
SC
P
R
,
S
C
MR
an
d
S
B
R
r
eg
is
ter
s
.
R
ec
o
n
f
ig
u
r
atio
n
is
also
p
er
f
o
r
m
ed
f
o
r
g
r
o
u
p
r
h
o
m
b
ic
to
ad
d
c
o
n
n
ec
tio
n
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Reco
n
f
ig
u
r
a
b
le
&
E
m
b
ed
d
ed
Sy
s
t
I
SS
N:
2089
-
4
8
6
4
C
o
s
t
-
efficien
t reco
n
fig
u
r
a
b
le
g
eo
metrica
l b
u
s
in
terco
n
n
ec
ti
o
n
s
ystem
…
(
Tir
u
ma
le
R
a
mes
h
)
81
Fig
u
r
e
7
.
Sy
s
tem
r
ec
o
n
f
ig
u
r
ati
o
n
co
n
t
r
o
ls
r
eg
is
ter
(
R
C
R
)
T
ab
le
2
g
iv
es a
n
illu
s
tr
ated
R
C
R
f
o
r
b
u
s
“2
”
f
o
r
GR
2
an
d
Q
R
.
T
ab
le
2
.
Sy
s
tem
r
ec
o
n
f
ig
u
r
atio
n
illu
s
tr
atio
n
f
o
r
n
=m
=
8
C
o
n
f
i
g
u
r
a
t
i
o
n
B
u
s
k
S
B
R
(
k
)
S
C
P
R
(
i
,
k
)
S
C
M
R
(
j
,
k
)
G
r
o
u
p
R
h
o
m
b
i
c
2
2
[
1
1
1
1
]
[
1
1
1
1
1
1
1
1
]
[
0
1
1
1
0
0
0
0
]
Q
u
a
d
r
a
n
t
R
h
o
m
b
i
c
2
[
1
1
1
1
]
[
1
1
1
1
1
1
1
1
]
[
0
1
1
1
0
1
1
1
]
3.
G
E
O
M
E
T
R
I
CA
L
B
U
S
I
N
T
E
RCO
NN
E
C
T
I
O
N
VA
L
I
D
AT
I
O
N
AN
D
ARB
I
T
RAT
I
O
N
W
h
en
r
an
d
o
m
b
u
s
is
ass
ig
n
ed
f
o
r
g
e
o
m
etr
ical
b
u
s
co
n
f
ig
u
r
atio
n
s
,
s
o
m
e
m
em
o
r
y
r
eq
u
es
ts
m
ay
n
o
t
co
m
p
lete
in
th
e
c
u
r
r
en
t
m
em
o
r
y
cy
cle
d
u
e
to
n
o
b
u
s
co
n
n
ec
t
io
n
to
th
e
m
e
m
o
r
y
; a
s
a
r
esu
lt,
it r
eq
u
ir
es sp
ec
if
ic
b
u
s
ar
b
it
r
atio
n
.
Fo
r
ass
ig
n
in
g
d
is
tin
ct
b
u
s
es
to
m
em
o
r
y
-
m
o
d
u
les,
we
r
e
q
u
ir
e
t
h
at
th
e
n
u
m
b
er
o
f
m
em
o
r
y
-
m
o
d
u
les
co
n
n
ec
ted
t
o
ea
ch
b
u
s
f
o
r
r
h
o
m
b
ic
co
n
n
ec
tio
n
is
eq
u
al
to
(
−
+
1
)
[
3
1
]
.
T
h
is
is
th
e
lo
we
r
b
o
u
n
d
co
n
d
itio
n
.
W
e
v
alid
ate
th
is
th
eo
r
em
f
o
r
o
u
r
p
r
o
p
o
s
ed
g
r
o
u
p
an
d
q
u
ad
r
an
t
co
n
f
ig
u
r
atio
n
s
u
s
in
g
th
e
c
o
r
o
llar
ies
in
th
e
f
o
llo
win
g
s
ec
tio
n
.
3
.
1
.
G
eo
m
e
t
rica
l bus
inte
rc
o
nn
ec
t
io
n c
o
nfig
ura
t
io
n v
a
lid
a
t
io
n
L
et
(
)
b
e
th
e
n
u
m
b
er
o
f
s
u
cc
es
s
f
u
l
d
is
tin
ct
b
u
s
c
o
n
n
ec
tio
n
s
i
n
ea
ch
g
r
o
u
p
i
f
o
r
=
1
,
.
.
wh
er
e
=
2
4
o
r
in
ea
c
h
q
u
a
d
r
an
t
i
f
o
r
=
1
,
…
4
.
−
C
o
r
o
lla
r
y
1
:
Fo
r
g
r
o
u
p
r
h
o
m
b
ic
co
n
f
i
g
u
r
atio
n
s
,
i
f
a
b
u
s
is
co
n
n
ec
te
d
to
less
th
an
−
+
1
m
em
o
r
y
-
m
o
d
u
les,
th
en
we
ca
n
n
o
t a
s
s
ig
n
b
d
is
tin
ct
b
u
s
es; in
th
is
ca
s
e,
ad
d
itio
n
al
co
n
n
ec
tio
n
s
n
e
ed
e
d
.
P
r
o
o
f
:
Fo
r
r
eq
u
ests
,
g
r
o
u
p
r
h
o
m
b
ic
s
atis
f
ies
th
e
lo
wer
b
o
u
n
d
c
o
n
d
itio
n
lo
ca
lly
(
with
in
g
r
o
u
p
)
.
Ho
wev
er
,
as
n
o
t
all
m
e
m
o
r
y
-
m
o
d
u
les
co
n
n
ec
tio
n
s
e
x
is
t
in
ea
ch
g
r
o
u
p
,
∑
(
)
=
1
<
.
Hen
ce
,
th
e
l
o
wer
b
o
u
n
d
co
n
d
itio
n
f
o
r
th
e
o
v
er
all
in
ter
co
n
n
ec
tio
n
will
n
o
t
b
e
s
atis
f
ied
an
d
r
eq
u
ir
e
at
m
o
s
t
ad
d
itio
n
al
co
n
n
ec
tio
n
s
.
−
C
o
r
o
lla
r
y
2:
L
et
=
{
1
,
2
,
…
/
2
}
an
d
B
=
{
(
2
)
+1
,
…
}.
Fo
r
b
m
em
o
r
y
r
e
q
u
ests
,
if
2
m
em
o
r
y
r
eq
u
ests
∈
an
d
2
m
em
o
r
y
r
eq
u
ests
∈
,
th
en
d
is
tin
ct
b
u
s
es
ca
n
b
e
ass
ig
n
ed
f
o
r
GR
2
.
T
h
is
ca
n
b
e
e
x
ten
d
ed
to
GR
4
with
s
ets
,
,
,
o
f
m
em
o
r
y
r
e
q
u
est ea
ch
m
ap
p
e
d
to
ea
ch
g
r
o
u
p
.
P
r
o
o
f:
I
n
g
r
o
u
p
r
h
o
m
b
ic,
with
g
r
o
u
p
s
,
/
r
h
o
m
b
ic
co
n
n
ec
tio
n
s
ex
is
t
givin
g
∑
(
)
=
1
=
g
.
b
/
g
=
.
I
n
th
is
ca
s
e,
th
e
lo
wer
b
o
u
n
d
i
s
s
atis
f
ied
in
ea
ch
g
r
o
u
p
an
d
d
is
tin
ct
b
u
s
es c
an
b
e
ass
ig
n
ed
.
−
C
o
r
o
lla
r
y
3:
Fo
r
q
u
a
d
r
an
t
r
h
o
m
b
ic,
i
f
e
v
er
y
b
u
s
is
co
n
n
e
cted
to
−
+
1
,
th
en
th
er
e
ex
is
ts
a
b
u
s
ar
b
itra
tio
n
alg
o
r
ith
m
th
at
ca
n
ass
ig
n
d
is
tin
ct
b
u
s
es f
o
r
b
m
e
m
o
r
y
-
m
o
d
u
les.
P
r
o
o
f
:
Fo
r
an
y
b
m
e
m
o
r
y
r
eq
u
ests
,
ea
ch
q
u
ad
r
an
t
is
co
n
n
e
cted
in
r
h
o
m
b
ic
a
n
d
s
atis
f
ies
th
e
lo
wer
b
o
u
n
d
[
3
1
]
with
in
ea
ch
q
u
ad
r
a
n
t.
I
n
th
is
ca
s
e,
(
)
tak
es
co
m
b
in
atio
n
g
iv
in
g
∑
(
)
4
=
1
=
.
Hen
ce
,
d
is
tin
ct
b
u
s
es
ca
n
b
e
ass
ig
n
ed
.
−
C
o
r
o
lla
r
y
4
:
L
et
=
{1
,
2
,
…
/
2
}
an
d
B
=
{
/
2
+1
,
…
}.
Fo
r
b
m
em
o
r
y
r
e
q
u
ests
,
if
2
m
em
o
r
y
r
eq
u
ests
∈
an
d
2
r
eq
u
ests
∈
,
th
en
d
is
tin
ct
b
u
s
es c
an
b
e
ass
ig
n
ed
in
q
u
ad
r
an
t r
h
o
m
b
ic.
P
r
o
o
f
:
Fo
r
q
u
ad
r
atic
r
h
o
m
b
ic
(
s
ee
Fig
u
r
e
5
)
,
th
e
lo
wer
b
o
u
n
d
co
n
d
itio
n
is
a
p
p
lied
t
o
(
/
2
−
/
2
+
1
)
m
em
o
r
y
-
m
o
d
u
les
co
n
n
ec
ted
t
o
ea
ch
q
u
ad
r
an
t.
T
h
ese
co
n
n
e
ctio
n
s
ar
e
co
n
tr
ib
u
ted
f
r
o
m
o
n
e
q
u
ad
r
an
t
ea
c
h
f
r
o
m
lef
t
h
al
f
q
u
ad
r
a
n
t
(
/
2
co
n
n
ec
tio
n
s
)
an
d
r
ig
h
t
h
alf
q
u
ad
r
a
n
ts
(
/
2
co
n
n
ec
tio
n
s
)
.
I
n
t
o
tal,
we
ca
n
ass
ig
n
d
is
tin
ct
b
u
s
es to
m
em
o
r
y
-
m
o
d
u
les.
3
.
2
.
G
eo
m
e
t
rica
l bus
a
rbit
ra
t
io
n a
lg
o
ri
t
hm
f
o
r
hiera
rc
hica
l a
nd
qu
a
dra
t
ic
rho
m
bic
L
et
b
e
th
e
s
et
o
f
m
em
o
r
y
r
eq
u
ests
s
o
r
ted
in
ascen
d
in
g
o
r
d
e
r
an
d
is
th
e
m
em
o
r
y
-
m
o
d
u
le
i
n
.
L
et
=
[
1
,
…
]
b
e
th
e
s
tatu
s
o
f
th
e
b
u
s
ass
ig
n
ed
with
[
]
=
0
an
d
=
0
in
itially
.
L
et
b
e
th
e
n
u
m
b
e
r
o
f
s
u
cc
ess
f
u
l
m
e
m
o
r
y
-
b
u
s
co
n
n
ec
ti
o
n
with
=
0
in
tially
.
L
et
[
,
]
b
e
th
e
co
n
n
ec
tio
n
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
9
-
4
8
6
4
I
n
t J Reco
n
f
ig
u
r
a
b
le
&
E
m
b
ed
d
ed
Sy
s
t,
Vo
l.
10
,
No
.
2
,
J
u
ly
2
0
2
1
:
7
7
–
89
82
f
r
o
m
to
.
T
h
e
alg
o
r
ith
m
s
ea
r
ch
es
f
o
r
a
b
u
s
in
o
r
d
er
f
o
r
e
v
er
y
M
in
s
et
A
an
d
g
r
a
n
ts
t
h
e
b
u
s
if
it
is
co
n
n
ec
ted
t
o
M.
I
f
th
e
b
u
s
>
,
th
e
n
th
e
n
e
x
t b
u
s
is
u
p
d
ate
d
to
“
1
”
an
d
th
e
s
ea
r
c
h
co
n
tin
u
es.
F
o
r
h
ier
ar
ch
ical
r
h
o
m
b
ic,
we
ar
b
itra
te
co
r
e
b
u
s
as:
C
o
r
e
bus
←
⌈
2
⌉
,
ex
am
p
le
co
r
e
b
u
s
=
1
if
co
r
e=
1
T
h
e
m
em
o
r
y
b
u
s
ar
b
itra
tio
n
g
i
v
en
as:
f
o
r
each
M
do
:
u
n
til
[
,
]
=
1
[
]
=
0
=
+
1
=
+
1
[
]
←
1
if
>
t
h
e
n
=
−
3
.
3
.
G
eo
m
e
t
rica
l bus
a
rbit
ra
t
io
n a
lg
o
ri
t
hm
f
o
r
g
ro
up
rho
m
bic
L
et
b
e
th
e
s
et
o
f
r
eq
u
ests
an
d
is
th
e
m
em
o
r
y
-
m
o
d
u
le
i
n
s
et
an
d
=
0
in
itially
.
L
e
t
b
e
t
h
e
n
u
m
b
e
r
o
f
s
u
cc
ess
f
u
l
m
em
o
r
y
-
to
-
b
u
s
co
n
n
ec
tio
n
s
with
=
0
in
itially
.
L
et
b
e
th
e
co
n
n
ec
tiv
ity
d
ef
in
ed
as
b
ef
o
r
e.
T
h
e
alg
o
r
ith
m
s
ea
r
c
h
es
a
b
u
s
f
o
r
e
v
er
y
in
s
et
an
d
g
r
an
ts
th
e
b
u
s
if
is
co
n
n
ec
ted
t
o
M.
I
f
th
e
b
u
s
>
,
th
en
th
e
n
e
x
t b
u
s
is
u
p
d
ated
to
b
u
s
“1
”
an
d
t
h
e
s
ea
r
ch
co
n
tin
u
e
s
.
T
h
e
m
em
o
r
y
b
u
s
ar
b
itra
tio
n
g
i
v
en
as:
fo
r
each
M
do
if
M
≤
:
if
[
,
]
=
1
a
n
d
[
]
=0
=
+
1
;
[
]
]
←
1
=
+
1
if
M
>
:
if
b
u
s>
b
bu
s
=
b
u
s
-
b
if
[
,
]
=
1
a
n
d
[
]
=0
=
+
1
;
[
]
]
←
1
=
+
1
3
.
4
.
G
eo
m
e
t
rica
l bus
a
rbit
ra
t
io
n si
m
ula
t
io
n
W
e
co
n
d
u
cted
ex
te
n
s
iv
e
g
eo
m
etr
ical
b
u
s
ar
b
itra
tio
n
s
im
u
latio
n
f
o
r
all
s
y
s
tem
s
izes
an
d
co
n
f
ig
u
r
atio
n
s
to
v
er
if
y
t
h
e
al
g
o
r
ith
m
.
W
e
s
h
o
w
th
e
illu
s
tr
ated
ass
ig
n
ed
b
u
s
es
to
m
e
m
o
r
y
(
s
h
ad
ed
)
in
Fig
u
r
es
8
th
r
o
u
g
h
9
.
−
S
imu
la
tio
n
1
: Q
u
ad
r
atic
R
h
o
m
b
ic:
Me
m
o
r
y
R
eq
u
est=
{
3
,
4
,
5
,
8
}
.
−
S
imu
la
tio
n
2
:
Gr
o
u
p
R
h
o
m
b
ic
2
: N
o
n
-
Fav
o
r
ab
le
R
eq
u
est =
{
1
,
5
,
6
,
7
}
−
S
imu
la
tio
n
3
:
Gr
o
u
p
R
h
o
m
b
ic
4
:
Fav
o
r
ab
le
R
eq
u
est= {
M2
,
M3
,
M5
,
M8
}
Fig
u
r
e
8
s
h
o
ws th
e
b
u
s
ass
ig
n
m
en
t,
Fig
u
r
e
9
s
h
o
ws th
e
b
u
s
ass
ig
n
m
en
t (
co
lo
r
s
h
ad
e
d
)
f
o
r
s
im
u
latio
n
2
in
d
icatin
g
th
at
ad
d
ed
c
o
n
n
ec
tio
n
R
o
n
r
ec
o
n
f
i
g
u
r
atio
n
i
s
r
eq
u
ir
ed
t
o
g
r
an
t
b
u
s
es
to
m
em
o
r
y
r
eq
u
est,
Simu
latio
n
3
is
s
h
o
w
n
in
Fig
u
r
e
9
(
c
o
lo
r
b
o
ld
e
d
)
f
o
r
f
av
o
r
ab
le
m
e
m
o
r
y
r
eq
u
ests
r
eq
u
ir
in
g
n
o
ad
d
itio
n
al
co
n
n
ec
tio
n
s
an
d
b
u
s
es c
an
b
e
ass
ig
n
ed
.
Fig
u
r
e
8
.
B
u
s
ass
ig
n
m
en
t f
o
r
q
u
ad
r
atic
r
h
o
m
b
ic
f
o
r
n
=m
=8
Fig
u
r
e
9
.
B
u
s
aa
s
s
ig
n
m
en
t illu
s
tr
atio
n
f
o
r
q
u
ad
r
atic
r
h
o
m
b
ic
f
o
r
n
=m=
8
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Reco
n
f
ig
u
r
a
b
le
&
E
m
b
ed
d
ed
Sy
s
t
I
SS
N:
2089
-
4
8
6
4
C
o
s
t
-
efficien
t reco
n
fig
u
r
a
b
le
g
eo
metrica
l b
u
s
in
terco
n
n
ec
ti
o
n
s
ystem
…
(
Tir
u
ma
le
R
a
mes
h
)
83
3
.
5
.
M
emo
ry
ba
nd
width
Fro
m
th
e
m
u
ltip
le
-
b
u
s
b
an
d
wi
d
th
an
aly
s
is
[
3
0
]
,
with
r
an
d
o
m
b
u
s
ar
b
itra
tio
n
,
t
h
e
m
em
o
r
y
b
an
d
wid
th
wi
th
r
e
duc
e
d
n
um
b
e
r
of
b
us
e
s
(
b
=
m
/
2
)
a
n
d
c
o
mpl
e
te
b
us
c
o
n
n
e
c
tion
s
is give
n
b
y:
=
−
(
2
)
T
h
e
f
ir
s
t
ter
m
in
(
2
)
is
th
e
cr
o
s
s
b
ar
b
an
d
wid
th
,
th
e
s
ec
o
n
d
ter
m
is
th
e
r
ed
u
ctio
n
in
b
a
n
d
wid
th
with
d
u
e
to
r
ed
u
ce
n
u
m
b
er
o
f
b
u
s
es.
As
th
e
n
u
m
b
er
o
f
b
u
s
co
n
n
ec
tio
n
s
r
ed
u
ce
d
f
o
r
g
eo
m
etr
ical
b
u
s
co
n
f
ig
u
r
atio
n
s
,
th
e
b
an
d
wid
th
in
(
2
)
is
f
u
r
th
er
r
ed
u
ce
d
b
y:
1
/
∑
(
−
=
1
)
(
3
)
W
h
er
e
is
th
e
n
u
m
b
er
o
f
m
e
m
o
r
y
-
m
o
d
u
les
co
n
n
ec
ted
t
o
b
u
s
an
d
(
−
)
is
th
e
n
u
m
b
er
o
f
m
em
o
r
y
-
m
o
d
u
les
n
o
t
co
n
n
ec
t
ed
to
b
u
s
i
.
A
s
eq
u
al
n
u
m
b
e
r
o
f
m
em
o
r
y
-
m
o
d
u
les
ar
e
c
o
n
n
ec
ted
to
ea
ch
b
u
s
in
a
r
h
o
m
b
ic
to
p
o
lo
g
y
[
2
9
]
,
th
e
s
ec
o
n
d
ter
m
in
eq
u
atio
n
(
4
)
g
i
v
es
th
e
av
er
ag
e
n
u
m
b
er
o
f
m
e
m
o
r
y
-
m
o
d
u
les
n
o
t
co
n
n
ec
ted
t
o
an
y
b
u
s
.
=
−
(
−
)
/
m
(
4
)
w
h
er
e
is
th
e
m
em
o
r
y
b
a
n
d
wi
d
th
u
s
in
g
g
eo
m
etr
ical
b
u
s
co
n
f
ig
u
r
atio
n
with
r
an
d
o
m
b
u
s
ar
b
itra
tio
n
.
W
e
an
aly
tically
d
er
iv
e
th
e
r
e
d
u
ctio
n
in
b
an
d
wid
th
in
(
4
)
u
s
i
n
g
lo
wer
b
o
u
n
d
co
n
d
itio
n
(
−
+
1
)
[
3
1
]
.
−
C
o
r
o
lla
r
y
5
:
W
h
en
a
b
u
s
i
s
ar
b
itra
ted
r
an
d
o
m
ly
f
o
r
h
ier
ar
ch
ical
an
d
q
u
a
d
r
an
t
g
e
o
m
etr
ical
b
u
s
co
n
f
ig
u
r
atio
n
s
,
th
e
r
e
d
u
ctio
n
in
b
an
d
wid
t
h
is
eq
u
al
to
(
−
1
)
/
2
f
o
r
1
n
u
m
b
er
o
f
m
em
o
r
y
-
m
o
d
u
les n
o
t
co
n
n
ec
ted
t
o
a
b
u
s
.
P
r
o
o
f
:
Fo
r
r
h
o
m
b
ic
b
ased
co
n
n
ec
tio
n
s
,
we
r
eq
u
ir
e
1
≤
−
(
−
+
1
)
=
1
≤
(
−
1
)
.
T
h
is
s
ets
a
th
r
esh
o
ld
f
o
r
1
f
o
r
r
e
d
u
cto
n
i
n
b
an
d
wid
th
to
−
1
.
T
h
e
av
e
r
ag
e
n
u
m
b
er
o
f
m
e
m
o
r
y
-
m
o
d
u
le
s
n
o
t
co
n
n
ec
ted
to
a
b
u
s
is
th
en
eq
u
al
to
(
−
1
)
/
.
T
h
e
av
er
ag
e
n
u
m
b
er
o
f
m
em
o
r
y
-
m
o
d
u
les
n
o
t
co
n
n
ec
t
ed
to
an
y
b
u
s
is
eq
u
al
to
(
−
1
)
/
wh
ich
r
ed
u
ce
s
to
(
−
1
)
/
2
f
o
r
=
2
.
T
h
e
r
ed
u
ctio
n
(
−
1
)
/2
is
also
eq
u
al
to
s
ec
o
n
d
ter
m
in
(
4
)
.
I
n
g
e
n
er
al,
m
em
o
r
y
b
an
d
wid
th
wit
h
r
an
d
o
m
b
u
s
ass
ig
n
m
en
t f
o
r
n
u
m
b
er
o
f
g
r
o
u
p
s
is
g
iv
en
b
y
:
=
−
(
−
+
−
1
)
/
2
(
5
)
we
s
ee
th
at
th
e
r
ed
u
ctio
n
in
b
a
n
d
wid
th
in
cr
ea
s
es with
.
Fo
r
h
ier
ar
ch
ical
an
d
q
u
ad
r
a
n
t r
h
o
m
b
ic,
we
ca
n
ap
p
l
y
=1
as a
s
p
ec
ial
ca
s
e
f
o
r
eq
u
atio
n
(
5
)
wh
ich
y
ield
s
th
e
r
ed
u
ctio
n
o
f
(
−
1
)
/
2
as st
ated
in
co
r
o
llar
y
5
.
−
C
o
r
o
lla
r
y
6
:
Fo
r
h
ier
ar
ch
ical
an
d
q
u
ad
r
a
n
t
co
n
f
i
g
u
r
ati
o
n
s
,
with
g
eo
m
etr
ical
b
u
s
ar
b
itra
tio
n
,
th
e
r
ed
u
ctio
n
i
n
b
an
d
wid
th
in
(
5
)
i
s
n
u
llified
.
P
r
o
o
f
:
Fo
r
2
m
em
o
r
y
-
m
o
d
u
les
co
n
n
ec
ted
t
o
ea
ch
b
u
s
;
it
r
eq
u
ir
es
2
≥
(
−
+
1
)
f
o
r
ass
ig
n
in
g
d
is
tin
ct
b
u
s
es
to
me
mory
-
m
o
d
u
les.
Fo
r
m
=2
b
g
iv
es
2=
b
+
1
.
As
2
>
-
1
,
th
e
r
ed
u
ctio
n
in
(
5
)
i
s
n
u
llified
g
iv
in
g
ef
f
ec
ti
v
e
b
an
d
wid
th
o
f
.
H
o
wev
er
,
f
o
r
g
r
o
u
p
r
h
o
m
b
ic,
th
er
e
s
till
ex
is
ts
s
o
m
e
s
m
all
r
ed
u
ctio
n
in
b
an
d
wid
th
f
r
o
m
b
as
2
<
−
1
.
B
u
t,
if
th
er
e
is
a
f
av
o
r
ab
le
m
em
o
r
y
r
e
q
u
est
p
atter
n
,
th
e
n
ea
ch
g
r
o
u
p
lo
c
ally
s
atis
f
ies
th
e
lo
wer
b
o
u
n
d
co
n
d
itio
n
[
3
1
]
y
ield
in
g
2
>
−
1
th
u
s
ass
ig
n
in
g
d
is
tin
ct
b
u
s
es
to
m
em
o
r
y
-
m
o
d
u
les.
Fo
r
co
m
p
le
te
b
u
s
co
n
n
ec
tio
n
s
,
we
ca
n
d
ed
u
ce
2
=
2
.
As
2
>
-
1
,
it
n
u
llifies
th
e
r
ed
u
ctio
n
in
b
a
n
d
wid
th
v
alid
atin
g
th
at
th
e
b
an
d
wid
th
f
o
r
m
u
ltip
le
b
u
s
s
y
s
tem
[
3
0
]
with
co
m
p
lete
b
u
s
c
o
n
n
ec
tio
n
s
is
g
iv
e
n
b
y
(
2
)
.
3
.
6
.
B
us
lo
a
d a
nd
bu
s
f
a
ult
t
o
lera
nce
B
u
s
lo
ad
is
th
e
n
u
m
b
er
o
f
m
e
m
o
r
y
-
m
o
d
u
les
co
n
n
ec
ted
to
e
ac
h
b
u
s
.
As
we
in
c
r
ea
s
e
th
e
b
u
s
lo
ad
,
th
e
ca
p
ac
itiv
e
lo
ad
in
g
o
n
th
e
b
u
s
i
n
cr
ea
s
es;
a
p
o
i
n
t
is
ev
e
n
tu
ally
r
ea
ch
ed
wh
e
n
s
p
ee
d
u
p
with
m
u
ltip
le
p
r
o
ce
s
s
o
r
s
g
ets
s
atu
r
ated
.
Me
m
o
r
y
lo
a
d
i
s
th
e
n
u
m
b
er
o
f
b
u
s
es
co
n
n
ec
t
ed
to
a
m
em
o
r
y
-
m
o
d
u
le
a
n
d
d
ictates
th
e
b
u
s
f
au
lt
to
ler
an
ce
.
I
f
a
b
u
s
i
∈
f
ails
,
th
en
it f
o
r
ce
s
th
e
lo
wer
b
o
u
n
d
to
(
−
+
2
)
[
3
1
]
.
W
e
lo
o
k
at
two
s
p
ec
if
ic
b
u
s
f
au
lt c
o
n
d
itio
n
s
f
o
r
h
ier
ar
ch
ic
al
an
d
q
u
ad
r
an
t c
o
n
n
ec
tio
n
s
:
C
r
itica
l B
u
s
e
s
:
B
u
s
es
with
a
m
em
o
r
y
lo
a
d
o
f
“1
”.
B
u
s
“1
”
an
d
“
b
“a
r
e
cr
itical
b
u
s
es.
I
f
cr
itical
b
u
s
es
f
ail,
t
h
en
th
e
m
em
o
r
y
-
m
o
d
u
le
“1
”
o
r
“
m”
is
co
m
p
letely
d
is
co
n
n
ec
ted
.
T
h
e
r
em
ed
y
is
to
p
r
o
v
id
e
an
ad
d
itio
n
al
co
n
n
ec
tio
n
R
to
cr
itical
b
u
s
es.
Fig
u
r
e
1
0
s
h
o
ws
th
e
b
u
s
ass
ig
n
m
en
t
(
c
o
lo
r
s
h
ad
e
d
)
f
o
r
a
cr
itical
b
u
s
f
au
lt
with
a
m
em
o
r
y
r
e
q
u
est
{M
1
,
M5
,
M6
,
M7
}
f
o
r
q
u
ad
r
a
n
t
r
h
o
m
b
ic.
I
n
th
is
ca
s
e,
m
em
o
r
y
-
m
o
d
u
le
“1
”
is
n
o
t
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
9
-
4
8
6
4
I
n
t J Reco
n
f
ig
u
r
a
b
le
&
E
m
b
ed
d
ed
Sy
s
t,
Vo
l.
10
,
No
.
2
,
J
u
ly
2
0
2
1
:
7
7
–
89
84
co
m
p
letely
d
is
co
n
n
ec
ted
.
T
h
u
s
,
we
s
ee
th
at
q
u
ad
r
an
t
r
h
o
m
b
ic
is
in
h
er
en
tly
f
au
lt
to
ler
an
t
to
cr
itical
b
u
s
es
an
d
en
s
u
r
es a
s
s
ig
n
m
en
t o
f
−
1
b
u
s
es r
eq
u
ir
in
g
n
o
a
d
d
itio
n
al
co
n
n
ec
t
io
n
s
.
N
o
n
-
C
r
itica
l
B
u
s
es
:
B
u
s
e
s
wit
h
a
m
e
m
o
r
y
lo
ad
>
1
ar
e
n
o
n
-
c
r
itical
b
u
s
es.
W
h
en
th
er
e
is
a
f
a
u
lt
o
n
a
n
o
n
-
cr
itical
b
u
s
,
th
e
m
e
m
o
r
y
b
an
d
wid
th
is
d
eg
r
a
d
ed
to
(
−
1
)
;
b
u
t
n
o
m
em
o
r
y
-
m
o
d
u
le
is
d
is
co
n
n
ec
ted
an
d
alwa
y
s
a
b
u
s
co
n
n
ec
tio
n
e
x
is
ts
f
o
r
all
m
em
o
r
y
-
m
o
d
u
les.
Gr
o
u
p
r
h
o
m
b
ic
is
less
b
u
s
f
au
lt to
ler
an
t a
s
th
e
n
u
m
b
e
r
o
f
c
r
itical
b
u
s
es
in
cr
ea
s
e
with
th
e
n
u
m
b
er
o
f
g
r
o
u
p
s
.
W
e
n
o
te
th
at
=
2
.
No
n
-
cr
itical
b
u
s
f
au
lts
ca
n
d
is
c
o
n
n
ec
t
2
(
/
−
/
+
1
)
m
em
o
r
y
-
m
o
d
u
les.
Ho
wev
er
,
th
e
ad
d
ed
co
n
n
ec
tio
n
s
ca
n
s
u
s
tain
th
ese
f
au
lts
.
E
v
en
with
th
e
in
cr
ea
s
e
in
co
s
t
f
o
r
ad
d
ed
co
n
n
ec
tio
n
s
,
th
e
g
r
o
u
p
i
n
ter
co
n
n
ec
tio
n
is
s
till
co
s
t e
f
f
ec
tiv
e.
Fig
u
r
e
1
0
.
Qu
ad
r
a
n
t r
h
o
m
b
ic
with
b
u
s
f
au
lt
4.
P
E
RF
O
RM
A
NCE CH
A
RA
CT
E
RI
Z
AT
I
O
N
In
th
is
s
ec
tio
n
,
we
p
r
esen
t
th
e
r
esu
lts
o
f
o
u
r
s
y
s
tem
ch
ar
ac
t
er
izatio
n
in
ter
m
s
o
f
co
s
t
p
er
b
an
d
wid
th
,
co
s
t p
er
d
eg
r
ad
ed
b
an
d
wid
t
h
a
n
d
s
y
s
tem
th
r
o
u
g
h
p
u
t w
ith
b
u
s
ca
ch
e.
4
.
1
.
Co
s
t
per
m
emo
ry
ba
nd
width
T
ab
le
3
s
h
o
ws
th
e
m
em
o
r
y
b
an
d
wid
th
o
f
g
eo
m
etr
ical
b
u
s
in
ter
co
n
n
ec
tio
n
wh
er
e,
co
l
1
r
ep
r
esen
ts
b
an
d
wid
th
with
r
a
n
d
o
m
b
u
s
a
r
b
itra
tio
n
f
r
o
m
eq
.
(
4
)
.
Fo
r
g
r
o
u
p
r
h
o
m
b
ic,
c
o
l
2
r
e
p
r
esen
ts
b
an
d
wid
th
with
n
o
ad
d
ed
co
n
n
ec
tio
n
s
an
d
co
l
3
r
ep
r
esen
ts
b
an
d
wid
th
with
ad
d
ed
co
n
n
ec
tio
n
s
(
b
o
th
s
h
ad
ed
)
b
y
u
s
in
g
g
eo
m
etr
ical
b
u
s
ar
b
itra
tio
n
al
g
o
r
ith
m
in
s
ec
tio
n
s
3
.
3
.
W
e
d
eter
m
in
ed
th
e
av
e
r
ag
e
b
an
d
wi
d
th
f
r
o
m
o
v
er
1
0
0
iter
atio
n
s
o
f
r
an
d
o
m
m
em
o
r
y
r
eq
u
est.
W
e
o
b
tain
ed
3
0
to
5
0
%
r
ed
u
ctio
n
in
m
em
o
r
y
b
a
n
d
wid
th
with
g
r
o
u
p
r
h
o
m
b
ic
wh
e
n
alg
o
r
ith
m
3
.
3
is
ap
p
lied
with
o
u
t
an
y
a
d
d
ed
co
n
n
ec
tio
n
s
.
Ho
wev
er
,
as
we
ad
d
ed
b
u
s
co
n
n
ec
tio
n
s
,
th
e
b
a
n
d
wid
th
in
cr
ea
s
es
to
(
−
1
)
.
Fo
r
HR
an
d
QR
,
c
o
l
2
r
ep
r
esen
ts
b
an
d
wid
th
f
r
o
m
alg
o
r
ith
m
3
.
2
g
iv
i
n
g
th
e
s
am
e
b
an
d
wid
t
h
as with
co
m
p
lete
b
u
s
co
n
n
ec
tio
n
s
(
r
o
w
1
)
.
T
ab
le
3
.
Me
m
o
r
y
b
an
d
wid
th
In
t
e
r
c
o
n
n
e
c
t
i
o
n
S
y
st
e
m
S
i
z
e
16
32
64
1
2
8
M
u
l
t
i
p
l
e
-
B
u
s
[
3
0
]
7
.
9
1
5
.
9
3
1
.
9
6
3
.
9
G
r
o
u
p
R
h
o
m
b
i
c
2
2
.
4
5
.
6
7
.
4
4
.
4
1
3
.
4
1
5
.
2
8
.
5
2
8
.
5
3
0
.
7
1
6
.
5
5
8
.
8
6
1
.
7
G
r
o
u
p
R
h
o
m
b
i
c
4
1
.
5
4
7
2
.
4
11
15
4
.
5
26
31
8
.
5
57
64
H
i
e
r
a
r
c
h
i
c
a
l
4
.
4
7
.
9
8
.
5
1
5
.
9
1
6
.
5
3
1
.
9
3
2
.
5
6
3
.
9
Q
u
a
d
r
a
n
t
4
.
9
7
.
9
8
.
9
1
5
.
9
17
3
1
.
9
33
6
3
.
9
T
ab
le
4
s
h
o
ws
th
e
co
s
t
p
er
b
an
d
wid
th
f
o
r
all
s
y
s
tem
s
izes.
Fo
r
g
r
o
u
p
r
h
o
m
b
ic,
co
l
1
an
d
co
l
2
r
ep
r
esen
ts
th
e
r
esu
lt
with
n
o
ad
d
ed
co
n
n
ec
tio
n
s
an
d
ad
d
ed
co
n
n
ec
tio
n
s
(
b
o
th
s
h
o
wn
s
h
ad
ed
)
an
d
co
l
3
r
ep
r
esen
ts
co
s
t p
er
b
an
d
wid
th
f
o
r
f
av
o
r
ab
le
m
em
o
r
y
r
eq
u
est.
Fig
u
r
e
1
1
s
h
o
ws th
e
av
er
ag
e
co
s
t p
er
b
an
d
wid
th
ac
r
o
s
s
all
s
y
s
tem
s
ize
s
.
T
h
e
r
ed
u
ctio
n
in
av
er
a
g
e
co
s
t
p
er
b
an
d
wid
th
v
ar
ies
f
r
o
m
1
.
3
to
1
.
8
co
m
p
ar
ed
to
co
m
p
lete
b
u
s
co
n
n
ec
tio
n
s
.
T
ab
le
4
.
C
o
s
t p
er
b
an
d
wid
th
I
n
t
e
r
c
o
n
n
e
c
t
i
o
n
S
y
st
e
m Si
z
e
16
32
64
1
2
8
C
o
m
p
l
e
t
e
B
u
s
C
o
n
n
e
c
t
i
o
n
s
32
64
1
2
8
2
5
6
G
r
o
u
p
R
h
o
m
b
i
c
2
2
9
.
8
2
4
.
8
21
4
8
.
8
4
4
.
8
41
9
1
.
2
8
5
.
8
81
1
7
4
.
7
1
6
8
.
5
1
6
1
G
r
o
u
p
R
h
o
m
b
i
c
4
38
2
6
.
3
19
5
3
.
8
4
5
.
7
37
8
9
.
8
8
2
.
1
73
1
6
2
1
5
1
1
4
5
H
i
e
r
a
r
c
h
i
c
a
l
/
Q
u
a
d
r
a
n
t
25
25
49
97
1
9
3
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Reco
n
f
ig
u
r
a
b
le
&
E
m
b
ed
d
ed
Sy
s
t
I
SS
N:
2089
-
4
8
6
4
C
o
s
t
-
efficien
t reco
n
fig
u
r
a
b
le
g
eo
metrica
l b
u
s
in
terco
n
n
ec
ti
o
n
s
ystem
…
(
Tir
u
ma
le
R
a
mes
h
)
85
Fig
u
r
e
1
1
.
Av
er
ag
e
c
o
s
t p
er
b
an
d
wid
th
4.
2
.
Co
s
t
per
deg
r
a
ded ba
nd
width
W
e
r
an
ex
ten
s
iv
e
s
im
u
latio
n
f
o
r
d
eter
m
in
i
n
g
th
e
d
eg
r
ad
e
d
b
an
d
wid
th
o
n
a
s
in
g
le
b
u
s
f
a
u
lt.
T
ab
le
5
s
h
o
ws th
e
av
er
ag
e
c
o
s
t p
er
b
a
n
d
wid
th
(
c
o
l 1
)
a
n
d
c
o
s
t p
er
d
eg
r
ad
ed
b
an
d
wid
t
h
(
co
l
2
)
.
T
ab
le
5
.
C
o
s
t p
er
b
an
d
wid
th
c
o
m
p
ar
is
o
n
s
I
n
t
e
r
c
o
n
n
e
c
t
i
o
n
S
y
st
e
m Si
z
e
16
32
64
1
2
8
G
r
o
u
p
R
h
o
m
b
i
c
2
-
R
2
4
.
8
2
8
.
2
4
4
.
8
4
8
.
3
8
5
.
8
8
8
.
3
1
6
8
.
5
1
7
0
.
2
G
r
o
u
p
R
h
o
m
b
i
c
4
-
R
2
6
.
3
3
6
.
2
4
5
.
7
4
8
.
1
8
2
.
1
8
2
.
1
1
5
1
1
5
3
.
4
H
i
e
r
a
r
c
h
i
c
a
l
/
Q
u
a
d
r
a
n
t
25
2
8
.
6
49
5
2
.
3
97
1
0
0
.
1
1
9
3
1
9
6
.
1
Fig
u
r
e
1
2
s
h
o
ws
th
e
a
v
er
ag
e
i
n
cr
ea
s
e
in
co
s
t
p
e
r
b
a
n
d
wid
th
(
f
r
o
m
co
l
1
to
co
l
2
in
T
a
b
le
5
)
ac
r
o
s
s
all
s
y
s
tem
s
izes.
T
h
e
p
er
ce
n
ta
g
e
av
er
a
g
e
in
cr
ea
s
e
in
co
s
t
p
e
r
b
an
d
wid
th
d
u
e
to
b
an
d
wid
th
d
eg
r
a
d
atio
n
v
ar
ies
f
r
o
m
3
.
6
to
4
.
8
.
Fig
u
r
e
1
2
.
Av
er
ag
e
c
o
s
t p
e
r
b
an
d
wid
th
co
m
p
ar
is
o
n
s
4.
3
.
E
f
f
ec
t
i
v
e
s
y
s
t
em
t
hro
ug
hp
ut
wit
h b
us
ca
che
As
th
e
n
u
m
b
e
r
o
f
p
r
o
ce
s
s
o
r
c
o
r
es
o
n
a
ch
ip
m
u
ltip
r
o
ce
s
s
o
r
in
cr
ea
s
ed
,
th
er
e
is
alwa
y
s
a
ch
allen
g
e
to
p
r
o
v
id
e
ad
eq
u
ate
in
ter
co
n
n
ec
t
io
n
b
an
d
wid
th
.
Use
o
f
m
u
lti
-
lev
el
ca
ch
e
ca
n
i
n
cr
ea
s
e
th
e
s
y
s
tem
th
r
o
u
g
h
p
u
t.
Ho
wev
er
,
u
s
e
o
f
lar
g
e
n
u
m
b
e
r
o
f
f
ast
o
n
-
c
h
ip
p
r
iv
at
e
co
r
e
ca
ch
es
in
cr
ea
s
es
th
e
s
y
s
tem
c
o
s
t.
A
m
u
c
h
s
lo
wer
s
h
ar
ed
b
u
s
ca
ch
e
p
lace
d
o
n
e
v
er
y
b
u
s
lin
e
ca
n
o
p
tim
ize
th
e
o
v
er
all
s
y
s
tem
co
s
t
an
d
r
ed
u
ce
av
er
ag
e
m
em
o
r
y
ac
ce
s
s
tim
e
lead
in
g
to
lo
wer
c
lo
ck
s
p
er
in
s
tr
u
ctio
n
(
C
PI)
.
T
h
e
h
it r
atio
o
f
b
u
s
ca
ch
e
ca
n
b
e
g
iv
en
as:
ℎ
=
(
−
)
/
(
6
)
W
h
er
e
is
th
e
cr
o
s
s
b
ar
b
an
d
wid
th
[
3
0
]
an
d
is
th
e
b
an
d
wid
th
u
s
in
g
g
eo
m
etr
ical
b
u
s
(
T
ab
le
3
)
.
T
a
b
le
6
s
h
o
ws
th
e
ef
f
ec
tiv
e
s
y
s
tem
th
r
o
u
g
h
p
u
t
with
b
u
s
ca
ch
e
.
I
n
T
a
b
le
6
,
c
o
l
1
c
o
r
r
esp
o
n
d
to
th
r
o
u
g
h
p
u
t
u
s
in
g
ℎ
f
r
o
m
(
6
)
an
d
co
l
2
r
ep
r
esen
ts
th
r
o
u
g
h
p
u
t
with
ℎ
in
cr
ea
s
ed
b
y
1
5
%
f
r
o
m
co
l
1
.
As
o
b
s
er
v
ed
in
T
a
b
le
6
,
we
s
ee
an
in
cr
ea
s
e
in
ef
f
ec
tiv
e
th
r
o
u
g
h
p
u
t
o
f
t
h
e
s
y
s
tem
wh
en
b
u
s
ca
ch
e
h
it
r
atio
is
in
cr
ea
s
ed
b
y
1
5
%.
Fig
u
r
e
1
3
s
h
o
ws
th
e
av
er
ag
e
t
h
r
o
u
g
h
p
u
t
f
o
r
ea
c
h
s
y
s
tem
s
ize.
W
e
o
b
s
er
v
e
th
at
th
e
av
er
ag
e
th
r
o
u
g
h
p
u
t
is
h
ig
h
er
f
o
r
h
ier
ar
c
h
ical
an
d
q
u
ad
r
an
t
r
h
o
m
b
ic
co
m
p
ar
e
d
to
g
r
o
u
p
r
h
o
m
b
ic.
Ho
wev
e
r
,
as
th
e
s
y
s
tem
s
ize
in
cr
ea
s
es,
th
e
d
if
f
er
en
ce
i
n
th
r
o
u
g
h
p
u
t
f
r
o
m
h
ier
ar
cc
h
ical
an
d
q
u
a
d
r
an
t
r
h
o
m
b
ic
t
o
g
r
o
u
p
r
h
o
m
b
ic
r
ed
u
ce
s
;
s
u
g
g
esti
n
g
an
ad
v
an
ta
g
e
f
o
r
u
s
in
g
b
u
s
ca
ch
e
f
o
r
g
r
o
u
p
r
h
o
m
b
ic
with
in
cr
ea
s
e
in
s
y
s
tem
s
ize.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
9
-
4
8
6
4
I
n
t J Reco
n
f
ig
u
r
a
b
le
&
E
m
b
ed
d
ed
Sy
s
t,
Vo
l.
10
,
No
.
2
,
J
u
ly
2
0
2
1
:
7
7
–
89
86
T
ab
le
6
.
Sy
s
tem
th
r
o
u
g
h
p
u
t w
ith
b
u
s
ca
ch
e
I
n
t
e
r
c
o
n
n
e
c
t
i
o
n
S
y
st
e
m Si
z
e
1
6
3
2
6
4
1
2
8
=
1
0
.
4
BW
cb
=
2
0
.
4
BW
cb
=
4
0
.
6
=
8
0
.
2
G
r
o
u
p
R
h
o
m
b
i
c
2
-
R
9
.
2
9
.
5
1
8
.
4
1
8
.
9
3
7
.
2
3
8
.
1
7
4
.
9
7
6
.
2
G
r
o
u
p
R
h
o
m
b
i
c
4
-
R
8
.
6
8
.
6
1
8
.
2
1
8
.
6
3
6
.
4
3
8
.
6
7
6
.
4
8
0
.
2
H
i
e
r
a
r
c
h
i
c
a
l
/
Q
u
a
d
r
a
n
t
9
.
8
1
0
.
1
1
9
.
5
1
9
.
9
3
8
.
8
3
9
.
8
7
7
.
6
7
8
.
9
Fig
u
r
e
1
3
.
Av
er
ag
e
s
y
s
tem
th
r
o
u
g
h
p
u
t w
ith
b
u
s
ca
ch
e
4.
4
.
E
s
t
im
a
t
ed
co
s
t
co
m
pa
riso
ns
t
o
No
C
CL
O
S net
wo
rk
W
e
co
m
p
ar
ed
g
e
n
er
ically
t
h
e
g
eo
m
etr
ical
b
u
s
co
n
f
ig
u
r
atio
n
s
co
s
t
with
a
cir
cu
it
s
witch
ed
No
C
[
1
6
]
b
ased
o
n
C
L
OS
n
etwo
r
k
[
3
3
]
u
s
in
g
th
e
s
am
e
s
y
s
tem
s
ize.
I
n
th
e
C
L
OS
b
ased
No
C
[
1
6
]
,
th
e
n
etwo
r
k
was
o
r
g
an
ized
as
3
lan
es
(
in
p
u
t,
m
id
d
le
an
d
o
u
tp
u
t)
co
n
s
is
tin
g
o
f
cr
o
s
s
b
ar
s
witch
es.
Fo
r
ex
am
p
le,
a
2
0
x
2
0
cir
cu
it
s
witch
ed
r
o
u
ter
h
as
f
iv
e
4
x
4
s
witch
es
in
th
e
in
p
u
t
lan
e,
f
o
u
r
5
x
5
s
witch
es
in
th
e
m
id
d
le
lan
e
an
d
f
iv
e
4
x
4
s
witch
es
in
th
e
o
u
tp
u
t
lan
e.
W
e
ass
u
m
ed
a
r
elativ
e
co
s
t
in
cr
ea
s
e
b
y
4
f
o
r
ea
c
h
s
witch
in
p
u
t/o
u
tp
u
t
in
C
L
OS
No
C
co
m
p
ar
ed
to
t
h
e
g
eo
m
etr
ical
b
u
s
co
n
n
ec
tio
n
s
witch
d
u
e
to
n
etwo
r
k
lay
e
r
s
an
d
b
u
f
f
er
s
.
W
e
u
s
ed
s
witch
in
p
u
t
u
n
it
co
s
t
i
n
cr
ea
s
e
f
o
r
r
eg
u
lar
cr
o
s
s
b
ar
b
y
2
co
m
p
ar
e
d
to
g
eo
m
etr
ical
b
u
s
co
n
n
ec
tio
n
s
witch
d
u
e
to
lar
g
er
s
witch
m
atr
ix
at
ea
ch
c
r
o
s
s
p
o
i
n
t.
T
ab
le
7
an
d
Fig
u
r
e
1
4
s
h
o
ws
th
e
esti
m
ated
co
s
t
co
m
p
ar
is
o
n
s
.
W
e
n
o
ticed
a
n
a
v
er
ag
e
r
e
d
u
ctio
n
b
y
2
in
esti
m
ated
co
s
t
o
f
th
e
g
eo
m
etr
ic
b
u
s
in
ter
co
n
n
ec
tio
n
co
m
p
ar
ed
to
C
L
OS b
ased
cir
c
u
it s
witch
ed
No
C
ac
r
o
s
s
all
co
n
f
ig
u
r
atio
n
s
.
T
ab
le
7
.
C
o
s
t c
o
m
p
a
r
is
o
n
s
with
cir
cu
it swit
ch
ed
No
C
[
1
6
]
S
i
z
e
HR
G
R
2
G
R
4
QR
N
o
C
N
o
C
C
o
n
f
i
g
u
r
a
t
i
o
n
16
2
0
0
1
6
8
1
5
2
2
0
8
7
6
8
(
4
,
4
x
4
,
)
,
(
4
,
4
x
4
)
,
(
4
,
4
x
4
)
32
7
8
4
6
5
6
5
9
2
8
0
0
2
0
4
8
(
8
,
4
x
4
)
,
(
4
,
8
x
8
)
,
(
8
,
4
x
4)
64
3
1
0
4
2
5
9
2
2
3
3
6
3
1
3
6
6
1
4
4
(
1
6
,
4
x
4
)
,
(
4
,
1
6
x
1
6
)
,
(
1
6
,
4
x
4
)
1
2
8
1
2
3
5
2
1
0
3
0
4
9
2
8
0
1
2
4
1
6
2
0
4
8
0
(
3
2
,
4
x
4
)
,
(
4
,
3
2
x
3
2
)
,
(
3
2
,
4
x
4
)
Fig
u
r
e
1
4
.
E
s
tim
ated
av
er
ag
e
co
s
t r
ed
u
ctio
n
f
ac
to
r
with
cir
c
u
it swit
ch
r
o
u
ter
No
C
4.
4
.
Su
mm
a
ry
o
f
re
s
ults
I
n
s
u
m
m
ar
y
,
o
u
r
r
esu
lts
ar
e
as
f
o
llo
ws:
−
Gr
o
u
p
r
h
o
m
b
ic
o
f
f
er
th
e
b
e
s
t
av
er
ag
e
co
s
t
s
av
in
g
s
(
3
6
%
to
4
2
%).
Fo
r
n
o
n
-
f
av
o
u
r
ab
le
m
em
o
r
y
r
eq
u
ests
ad
d
iti
o
n
al
co
n
n
ec
tio
n
s
ar
e
r
eq
u
ir
e
d
to
ac
h
iev
e
s
am
e
m
em
o
r
y
b
an
d
wid
th
o
f
.
E
v
en
b
y
ad
d
in
g
at
m
o
s
t
n
u
m
b
e
r
o
f
co
n
n
ec
tio
n
s
,
g
r
o
u
p
r
h
o
m
b
ic
s
till
o
f
f
e
r
s
th
e
b
est
a
v
er
ag
e
co
s
t
s
av
in
g
s
m
ak
in
g
it
a
g
o
o
d
c
h
o
ice
f
o
r
h
ig
h
er
s
y
s
tem
s
izes.
−
W
e
ac
h
iev
ed
r
ed
u
cti
o
n
o
f
1
.
5x
co
s
t p
er
m
em
o
r
y
b
an
d
wid
th
with
g
r
o
u
p
r
h
o
m
b
ic
ac
r
o
s
s
all
s
y
s
te
m
s
izes.
−
Qu
ad
r
atic
an
d
h
ier
ar
ch
ical
r
h
o
m
b
ic
ac
h
ie
v
es
th
e
s
am
e
m
em
o
r
y
b
an
d
wid
th
with
o
u
t
r
eq
u
ir
in
g
an
y
ad
d
itio
n
al
co
n
n
ec
tio
n
s
an
d
we
ac
h
iev
ed
r
e
d
u
ctio
n
o
f
1
.
3x
co
s
t p
e
r
m
em
o
r
y
b
a
n
d
wid
th
.
Evaluation Warning : The document was created with Spire.PDF for Python.