Internati
o
nal
Journal of Ele
c
trical
and Computer
Engineering
(IJE
CE)
V
o
l.
4, N
o
. 2
,
A
p
r
il
201
4, p
p
.
28
5
~
29
4
I
S
SN
: 208
8-8
7
0
8
2
85
Jo
urn
a
l
h
o
me
pa
ge
: h
ttp
://iaesjo
u
r
na
l.com/
o
n
lin
e/ind
e
x.ph
p
/
IJECE
Metric Suite to Evaluate Reusab
ility of Software Product Line
Mohammad Ali
Torkam
ani
R&D Department, Ir
anian Te
lecommunication Manufactur
ing Compan
y
,
Shiraz,
I
r
an
Article Info
A
B
STRAC
T
Article histo
r
y:
Received Nov 13, 2013
Rev
i
sed
Jan 19, 201
4
Accepte
d
Fe
b 4, 2014
Metrics hav
e
lo
ng been used
to
measur
e and
evaluate software p
r
oducts an
d
processes. Software product lin
e
architec
ture
is a
field
in which f
e
w metrics
have been app
l
i
e
d, a s
u
rpris
i
ng
fact
given th
e important role
of software
product line architecture in software
product line dev
e
lopment. Recen
tly
,
Some metrics have b
een d
e
v
e
lope
d
to assess software product lin
e
archi
t
ec
ture
. Th
es
e m
e
trics
are
us
eful
but have not been widely
used in
industr
y
.
In th
is paper, som
e
new
m
e
trics are prov
ided to
assess reusabilit
y
o
f
S
o
ftware produ
c
t
lin
e
arch
ite
ctur
e. Our
m
e
tri
c
s
ar
e ev
alu
a
ted
in
ac
tion.
Keyword:
Reu
s
ab
ility
Soft
ware
P
r
o
d
u
ct line
Soft
ware
m
e
trics
Copyright ©
201
4 Institut
e
o
f
Ad
vanced
Engin
eer
ing and S
c
i
e
nce.
All rights re
se
rve
d
.
Co
rresp
ond
i
ng
Autho
r
:
Mohammad Ali
Torkam
ani
R
&
D De
part
m
e
nt
, Ira
ni
an Tel
ecom
m
uni
cat
ion
Manufact
uring C
o
m
p
any,
Shiraz
, Ira
n
1.
INTRODUCTION
A so
ft
wa
re p
r
od
uct
l
i
n
e i
s
a set
of so
ft
wa
r
e
-i
nt
ensi
ve sy
s
t
em
s shari
ng a
com
m
on,
m
a
nage
d set
o
f
feature
s
that sa
tisfy the specific need
s
o
f
a
p
a
rt
i
c
ul
ar m
a
rke
t
segm
ent
or m
i
ssi
on a
n
d t
h
at
are de
vel
o
pe
d
fr
om
a comm
on set of core as
sets in a
pre
s
cribe
d
way [1
]. Me
trics are em
ployed
for estimating s
o
ft
ware
s a
nd
pr
ocesses
[
2]
. Avai
l
a
bl
e m
e
t
r
i
c
s i
n
soft
ware
engi
neeri
ng a
r
e i
n
s
u
f
f
i
c
i
e
nt
and e
v
e
n
are
di
ffi
c
u
l
t
y
appl
i
e
d f
o
r
est
i
m
a
ti
ng P
r
o
duct
Li
ne
A
r
c
h
i
t
ect
ure
(PL
A
).
PL
A
i
s
a
fi
el
d
wi
t
h
l
e
ss de
fi
ne
d m
e
t
r
i
c
s by
w
h
i
c
h i
t
i
s
estim
a
ted. In recent years, s
o
me
m
e
trics have bee
n
i
n
troduced
for estimating PL
A.
Although t
h
ese metrics
are ve
ry
use
f
u
l
, t
h
ey
ha
ve n
o
t
bee
n
ho
we
ver
,
wi
del
y
em
pl
oy
ed i
n
i
n
dust
r
i
e
s. F
o
r
t
h
i
s
, e
xpe
rt
s an
d R
&
D
depa
rt
m
e
nt
s shoul
d pay
m
o
re
at
t
e
nt
i
on t
o
t
h
e
m
e
t
r
i
c
s em
pl
oy
i
ng i
n
pr
o
duc
t
l
i
n
es. I
n
t
h
i
s
pape
r,
we i
n
t
r
o
duc
e
so
m
e
metrics
for esti
m
a
tin
g
reu
s
ab
ility in
so
ft
ware
p
r
od
u
c
t lin
es. The rest o
f
t
h
is p
a
p
e
r is stru
ctu
r
ed
as
fo
llows.
After ex
p
l
ain
i
n
g
t
h
e related
wo
rk
s i
n
th
e seco
nd
part, th
e m
e
tric
su
ite for ev
al
uatin
g
reu
s
ab
ilit
y in
so
f
t
w
a
r
e
pr
oduct lin
e is ex
p
l
ain
e
d
i
n
t
h
e th
i
r
d
p
a
r
t
. Th
en in th
e
f
our
th
p
a
r
t
, case st
ud
y w
i
ll b
e
ex
p
l
ain
e
d. Th
e
l
a
st
part
of
pa
p
e
r
has
been
al
l
o
cat
ed t
o
c
o
ncl
u
si
o
n
.
2.
RELATED WORKS
Mo
st o
f
in
itial wo
rk
on
so
ft
ware m
e
trics fo
cused
o
n
co
de
m
e
trics wh
ich
are d
e
riv
e
d
so
lely fro
m
sou
r
ce c
o
de o
f
apr
o
gram
, suc
h
as Li
nes
of
C
ode
, Ha
lstead
’s m
e
trics and
M
cCab
e’s cy
clo
m
a
tic co
m
p
lex
ity.
As t
h
e de
vel
o
p
m
ent
ofo
b
j
ect
-
o
ri
e
n
t
e
d t
ech
n
o
l
o
gy
, som
e
object
-o
ri
ent
e
d m
e
t
r
i
c
s havebe
en pr
o
p
o
s
ed
, such as
C
K
m
e
t
r
i
c
, and M
O
O
D
m
e
t
r
i
c
.Som
e
com
pone
nt
m
e
t
r
i
c
s al
so are pro
p
o
se
d t
o
m
easureco
m
p
l
e
xi
ty
,
custom
izability, and re
usa
b
i
lity of com
ponents
.
Existing
softwa
re m
e
trics are infle
x
ible and insufficient
form
easu
r
in
g PLA.
PLA repr
esen
t referen
c
e
arch
itectu
r
e o
f
p
r
od
u
c
t
li
n
e
m
e
m
b
ers. Variab
ility is b
a
sis for
i
m
p
l
e
m
en
tin
g
p
articu
l
arity o
f
p
r
od
u
c
t lin
e me
m
b
ers,
v
a
ri
abilit
y
m
e
trics areo
n
e
m
o
st i
m
p
o
r
tan
t
p
a
rt of
PLA
metrics. Variab
ility also
m
a
k
e
sPLA m
o
re co
m
p
lex
,
and
co
m
p
lex
ity
m
e
t
r
ics of PLA mu
stcon
s
id
er issu
es of
v
a
riab
ility. PLA will b
e
reu
s
ed
b
y
p
r
od
u
c
tl
in
e m
e
m
b
ers, reu
s
ab
ility als
o
sh
ou
ld
b
e
assessed
.
So
somen
e
w
metrics
m
e
th
od
s shou
ld
b
e
p
r
op
o
s
ed
to
measu
r
e qu
ality o
f
PLA[3
]
. So
m
e
refe
ren
c
es lik
e [3-10
]
h
a
ve
p
r
op
o
s
ed
so
m
e
m
e
trics fo
r measu
r
ing
qu
ality
in
so
ftware p
r
odu
ct lin
es. Th
e m
o
st i
m
p
o
r
tan
t
o
f
the
m
are:
Stru
ct
u
r
e Sim
i
l
a
rity Co
efficien
t (SSC), C
o
mp
on
en
t
Reuse
Rate (CRR), R
e
u
s
e Ben
e
fit R
a
te (RBR),
Prod
u
c
t
-
related
Reu
s
abilit
y (PrR), Size o
f
Co
mm
o
n
a
lity
(SOC) an
d Percen
t Reu
s
e (PR). Also
, [11
-
16
] h
a
v
e
prop
o
s
ed
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
IJEC
E V
o
l
.
4, No
. 2, A
p
ri
l
20
14
:
28
5 – 2
9
4
28
6
metrics for assessin
g
th
e featu
r
e m
o
d
e
l.
In
n
e
x
t
s
ectio
n we
will in
trod
u
c
e a
few metrics fo
r estimatin
g
reu
s
ab
ility in
so
ft
ware
produ
ct lin
es.
3.
METRI
C
S
U
I
TE FO
R EV
ALU
A
TIN
G
REUS
ABILI
T
Y I
N
S
O
FT
WA
RE P
R
O
D
U
C
T LI
NES
Th
e m
a
in
o
b
j
ect o
f
a product lin
e is reu
s
ab
ility [3
]. V
a
riou
s assets are b
e
ing
u
s
ed
i
n
softw
a
re
pr
o
duct
l
i
n
es.
These asset
s
h
a
ve di
f
f
ere
n
t
v
a
l
u
es. Al
s
o
, t
h
e val
u
es
of t
h
e
m
di
ffer f
r
o
m
the val
u
e o
f
t
h
e
pr
ofi
t
obt
ai
ne
d
by
or
gani
zat
i
o
n t
h
r
o
u
g
h
em
pl
oy
i
n
g
re
use
ap
pr
o
ach.
Al
t
h
o
u
g
h
t
h
e as
set
s
w
h
i
c
h are
b
e
i
n
g
use
d
i
n
pr
o
duct
l
i
n
es
h
a
ve di
ffe
rent
v
a
l
u
es, m
o
st
avai
l
a
bl
e
m
e
trics
lik
e SOC and
SSC [3
-5
] ho
wev
e
r, don
’t consid
er
th
e weigh
t
v
a
lu
es
o
f
th
ese assets. In
th
is
p
a
p
e
r
we
propo
se m
e
trics wh
ich
co
nsid
er
th
e weigh
t
v
a
lu
es
o
f
assets.
3.1.
Determining
the Weight Val
u
e of
Asse
ts
In
p
a
st years, t
h
e fo
cu
s
o
f
exp
e
rts
was
on
t
h
e reu
s
ab
ility o
f
fin
e
grain
assets lik
e reu
s
ab
ility in
co
d
e
lev
e
l. Du
e to
th
is app
r
o
a
ch
,
we h
a
v
e
seen
fewer su
ccesses in
reu
s
ab
ility
field
.
Curren
tly, th
e co
n
c
en
tratio
ns
have
bee
n
cha
nge
d t
o
war
d
s
coarse
grai
n asset
s
whi
c
h ar
e bei
ng
u
n
i
f
o
r
m
e
d by
soft
w
a
re arc
h
i
t
ect
ur
e. Thi
s
app
r
oach
has s
o
m
e
advant
a
g
e
s
:
a) t
h
e asset
s
wo
ul
d
be m
o
re
app
r
op
riate fo
r o
ffe
rin
g
in m
a
rket,
b)
it incr
eases
pr
o
duct
i
v
i
t
y
a
n
d
c) i
t
sa
ves
t
i
m
e
. [17]
M
o
r
e
ove
r,
t
h
e
SEI f
r
a
m
e
w
o
r
k
of p
r
od
uct lin
e [6
] con
s
id
er
s
pr
odu
ct
lin
e as an
attem
p
t fo
r em
p
l
o
y
in
g
strate
gi
c
pl
ans
fo
r c
o
ars
e
grai
n re
use.
For
t
h
is, larger grai
n assets are
m
o
re
val
u
a
b
l
e
f
o
r re
usi
n
g i
n
s
o
ft
w
a
re
pr
od
uct
l
i
n
e. I
n
or
de
r to
d
e
term
in
e th
e
weigh
t
v
a
l
u
e
o
f
assets
we sh
ou
l
d
conve
r
t assets
and artifacts t
o
a c
o
m
m
on m
easurem
ent
uni
t
s
u
ch
as "
L
i
n
e o
f
c
ode”
[1
0]
. I
f
t
h
e
n
u
m
ber of
code l
i
n
es
of t
h
e so
ft
wa
re as
set
s
i
s
not
avai
l
a
bl
e( l
i
k
e a si
t
u
at
i
on i
n
w
h
i
c
h an
or
ga
ni
zat
i
on has
p
u
rc
h
a
sed a
co
mmercial o
f
th
e sh
el
f (C
OTS)
) or it is d
i
fficu
lt to
u
s
to
co
nv
er
t no
n sof
t
w
a
r
e
assets to
th
e nu
m
b
er
of
co
de
l
i
n
es, we c
a
n
u
s
e an a
p
p
r
oach
i
n
o
r
de
r t
o
det
e
rm
i
n
e t
h
e wei
ght
val
u
e
o
f
as
set
s
. Su
p
p
o
s
e t
h
at
am
ong
di
ff
eren
t
asset
s
, t
h
e
ak
r
e
qui
res t
h
e
m
i
ni
m
u
m
effort
f
o
r
de
vel
o
pi
n
g
.
Thi
s
m
i
nim
u
m
eff
o
rt
i
s
sh
o
w
n
by
Ek
.
No
w,
we ca
n
calcu
late th
e
weig
h
t
v
a
lu
e of t
h
e ai asset t
h
rou
g
h
equ
a
tio
n (1
):
W
(1
)
It is clear th
at th
e weigh
t
v
a
lu
e o
f
t
h
e asset ak
will b
e
eq
u
a
l to
on
e. Th
e h
i
g
h
e
r lev
e
ls o
f
effort
require
d
for de
veloping an asset will
have
more costs. For this, in the eq
ua
tion (1) we can replace effort
level
by
de
vel
o
pm
ent
cost
. T
h
e
n
,
w
e
ha
ve:
W
(2
)
3.2.
Weight Perce
nt
of Re
usability
We can
im
p
r
ov
e th
e SSC metric b
y
ap
plyin
g
weigh
t
v
a
lu
es. As
our m
e
tric d
i
ffers fro
m
SC
C
form
u
l
a, we call it weig
h
t
p
e
rcen
t of
reu
s
ab
i
lity. Acco
rd
ing to
eq
u
a
tion
(3),
weigh
t
p
e
rcen
t of
reu
s
ab
ility is:
(the s
u
m
of c
o
m
m
on com
ponents
of
PL
A /the
sum
of all c
o
m
ponents
of product line
)
*100:
Wt
%
R
∑
∑
100
(3
)
In w
h
i
c
h k i
s
t
h
e n
u
m
b
er of
com
m
on co
m
pone
nt
s o
f
PLA
,
n i
s
t
h
e t
o
t
a
l
num
ber o
f
co
m
ponent
s o
f
pr
o
duct
l
i
ne,
W
i
s
t
h
e
wei
g
ht
val
u
e
o
f
t
h
e i
t
h c
o
m
m
on co
m
ponent
a
n
d
A
is
th
e weigh
t
v
a
lu
e
of
t
h
e j
t
h
com
pone
nt. Ac
cording to this form
ula,
the highe
r wei
ght va
lues of c
o
m
m
o
n
com
pone
nts
of PL
A will lead to
th
e h
i
g
h
e
r arch
itectu
r
al sim
i
l
a
rity o
f
t
h
e m
e
m
b
ers o
f
pr
odu
ct lin
e
wh
ich in
turn
will lead
to h
i
g
h
e
r
rates o
f
p
r
o
f
it
o
b
t
ai
n
i
ng
thro
ugh
em
p
l
o
y
in
g
reu
s
ab
ility ap
p
r
o
a
ch
.
If
we
sh
ow
t
h
e weigh
t
v
a
l
u
e of p
r
o
d
u
c
t
lin
e assets
as
W
s
p
l
,
we can
rewrite th
e eq
u
a
tion
(3
) as fo
llo
ws:
Wt
%
R
∑
w
100
(4
)
Also
,
we can
calcu
l
ate weigh
t
p
e
rcen
t
of
reusab
ility fo
r
p
r
od
u
c
t
lin
e
p
r
od
ucts th
rou
g
h
t
h
e fo
llo
wi
ng
equat
i
o
n:
Wt
%
R
∑
w
100
(5
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
Metric su
ite to
Eva
l
ua
te Reu
s
a
b
ility o
f
So
ftwa
r
e Produ
ct Li
n
e
(Moha
mmad
Ali To
rka
m
an
i)
28
7
In w
h
ich
Wt
%
R
is the weigh
t
p
e
rcen
t of reu
s
ab
ility o
f
t
h
e pro
duct P and
wj
is t
h
e
weigh
t
v
a
l
u
e o
f
the jt
h c
o
m
ponent t
h
at re
use
d
in the
product p.
Wp is
t
h
e
weight value
of
the assets of product
P whic
h
is
deri
ved
t
h
r
o
ug
h t
h
e
f
o
l
l
o
wi
n
g
eq
uat
i
on:
w
∑
w
(6
)
In wh
ich m
is t
h
e
n
u
m
b
e
r
o
f
th
e assets of th
e produ
ct P.
3
.
3
.
Av
erag
e
o
f
Reha
bilita
t
ion
If Ci be the average
of re
ha
bilitation of the
ith co
m
pone
nt in software
produ
ct line, the avera
g
e of
reh
a
b
ilitatio
n
o
f
who
l
e assets in
so
ftware p
r
odu
ct
lin
e (Ao
R
sp
l
)
wo
uld
b
e
d
e
ri
v
e
d fro
m
th
e fo
llo
wi
n
g
equat
i
o
n:
Ao
R
∑
C
(7
)
In
w
h
i
c
h
k i
s
t
h
e
num
ber
of
r
e
use
d
c
o
m
pon
ent
s
i
n
t
h
e c
o
m
m
on pa
rt
o
f
PLA a
n
d
n i
s
t
h
e
num
ber
of
wh
ol
e
com
p
o
n
e
nt
s i
n
t
h
e
co
m
m
on
part
o
f
PLA.
T
h
e
val
u
e o
f
C
wo
ul
d
be
one
i
f
t
h
e i
t
h
c
o
m
pone
nt
be
u
s
ed
as
Black Box. F
o
r
other re
usa
b
i
lity
m
e
thods
like
Whit B
o
x
approach, the
value
of Ci
is
obtaine
d though the
fo
llowing
equ
a
tio
n
:
C
1
(8
)
In
w
h
ich
M
is the
perce
n
t
of
changes
applying on
each com
ponent
for a
d
option a
n
d
re
usa
b
ility
pu
r
poses
.
Si
m
ilarly, we
can
calcu
late t
h
e av
erag
e of
reh
a
b
ilitatio
n
o
f
a g
i
v
e
n
prod
u
c
t throug
h
th
e fo
ll
o
w
i
ng
equat
i
o
n:
Ao
R
∑
Cp
(9
)
In
w
h
ich
Ao
R
is
th
e av
erag
e
of reh
a
b
ilitatio
n
o
f
th
e pro
d
u
c
t
p
an
d
Cp
i
s
the
ave
r
age
of
reha
bilitation of the ith c
o
m
pone
nt in
t
h
e product p. T
h
e
value of
Cp
is calcu
lated
sim
i
lar to
C
i.e. th
ro
ugh
t
h
e eq
uat
i
on
8.
In t
h
e e
quat
i
o
n (
9
),
k i
s
t
h
e
num
ber o
f
t
h
e
reuse
d
c
o
m
ponent
s i
n
t
h
e p
r
od
uct
p a
nd
n
i
s
t
h
e
total num
ber of com
pone
nts in the product p. If we
wish
to express
the
a
v
erage of
re
ha
bilitee
in perce
n
t, it is
ju
st
en
o
u
g
h
t
o
m
u
lt
i
p
l
y
t
h
e de
ri
ve
d
num
bers
fr
om
t
h
e equat
i
ons
(
7
)
a
n
d
(
9
)
by
10
0.
Exam
pl
e:
im
agi
n
e fi
ve c
o
m
p
o
n
ent
s
as C
1
t
o
C
5
whi
c
h
ha
ve
bee
n
re
used
i
n
a
s
o
ft
wa
re
p
r
od
uct
l
i
n
e
.
Tabl
e 1 s
h
o
w
s
t
h
e perce
n
t
o
f
chan
ges o
f
t
h
ese com
pone
nt
s. In t
h
i
s
t
a
bl
e,
W
B
st
a
nds
fo
r
Whi
t
e
B
ox a
nd B
B
stan
d
s
fo
r Black
Bo
x.
Tabl
e 1.
Th
e pe
r
c
e
n
t
o
f
ch
ang
e
f
o
r
ad
op
ting w
ith
n
e
w
ar
c
h
ite
c
t
u
r
e
Row
Na
m
e
type
percent
of
change for
adopt
i
on (
M
i
)
Ci
1
C1
WB
20%
0.
8
2
C2 BB
0%
1
3
C3
WB
50%
0.
5
4
C4 W
B
35%
0.
65
5
C5
WB
70%
0.
3
Th
e av
erag
e of reh
a
b
ilitatio
n
o
f
software
p
r
od
u
c
t lin
e is ex
pressed
as
fo
llows:
Ao
R
1
5
0.8
1
0
.5
0
.65
0.3
0
.
6
5
It cou
l
d b
e
said th
at th
e
av
erage of reh
a
b
ilitati
o
n
fo
r adop
ting
with
so
ft
ware produ
ct lin
e i
s
65
%.
3
.
4
.
Intro
ducing
So
me Metrics fo
r
Estima
ting
Reus
a
b
ility
B
a
sed
On
the
Mapping
of So
ftwa
re Pro
d
uct
Line as
Grap
h
Recently, some of researc
h
e
r
s like
Mr. Burger[5] ha
ve em
ployed the
theory of sets and
gra
ph for
m
odel
i
ng s
o
ft
ware
p
r
o
d
u
ct
l
i
n
e an
d
di
spl
a
y
i
ng t
h
e
rel
a
t
i
ons
hi
ps
o
f
t
h
e p
r
o
d
u
ct
s o
f
pr
o
duct
l
i
n
e
.
I
n
t
h
i
s
sectio
n
,
we introd
u
ce so
m
e
metrics fo
r software reu
s
ab
i
lity. Th
ese
m
e
trics h
a
v
e
b
e
en
ob
tain
ed
thro
ugh
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
IJEC
E V
o
l
.
4, No
. 2, A
p
ri
l
20
14
:
28
5 – 2
9
4
28
8
mapping s
o
ftware product line to a graph, t
h
at we ca
ll it Produ
ct-Asset
g
r
aph
(see
fig. 1
)
. Assu
m
e
th
at A is
th
e set
o
f
t
h
e assets of
o
u
r
p
r
od
u
c
t lin
e:
A
a
1
,
a
2
,…,a
|
|
In
th
is set, t
h
e n
u
m
b
e
r of th
e
me
m
b
ers o
f
the set A is sh
own
as
|
A
|
. Also, ass
u
m
e
that P is
the set
of
th
e pro
d
u
c
ts
o
f
th
e so
ft
ware produ
ct lin
e:
P
p
1
,
p
2
,…,p
|
|
Ag
ai
n
i
n
th
is set, th
e
nu
m
b
er
o
f
th
e m
e
m
b
ers of t
h
e set
P i
s
shown as
|
P
|
. E
ach ass
e
t can be use
d
i
n
ev
ery
p
r
od
uct. Assu
m
e
th
at th
e percen
t
o
f
ch
an
ges
ap
p
l
yin
g
for
ado
p
t
in
g
reu
s
ab
ility
in
d
i
fferen
t produ
cts
d
i
f
f
e
r
s
fr
o
m
as
set to
asset. Th
is i
m
p
lies th
a
t
th
e p
r
of
it o
b
t
ain
i
n
g
thro
ugh th
e r
e
u
s
ing
of assets in
p
r
odu
cts
wo
ul
d be
di
ffe
rent
. Ass
u
m
e
that
B
i
s
t
h
e ben
e
fi
t
obt
ai
ni
ng t
h
r
o
ug
h t
h
e r
e
u
s
i
ng
of asset
a
i
n
pr
o
duct
P
.
We defi
ne
t
h
e
wei
g
ht
ed
a
n
d d
i
rect
ed gra
p
h G
as f
o
l
l
o
w
s
:
G= (
V
,
E
)
VP
A
EP
A
B
B
∀
B
∶
i
∈
P
,
j
∈
A ,
B
C
D
C
R
B
,
C
D
,
C
R
∈
B
is th
e
b
e
nefit ob
tain
ing
t
h
ro
ugh
the reu
s
in
g
of asset
a
i
n
pr
o
duct
p
.
CD
is th
e co
st
of
devel
opi
ng
ass
e
t
aj i
n
t
h
e
p
r
od
uct
p
.
CR
is th
e
co
st
o
f
reu
s
ing th
e asset aj
i
n
th
e
p
r
od
u
c
t
p.
This gra
p
h
i
n
cl
ude
s t
h
e
co
upl
e
o
f
e
dges
l
i
k
e
e
,
p
1
,a
2
,B
1
2
∈E
.
Figure 1.
Product-asset gr
aph
We
will in
trodu
ce so
m
e
m
e
tri
c
s b
a
sed
on
th
i
s
graph
i
n
n
e
x
t
.
3
.
4
.
1
.
Ca
lculating
the
Benefi
t
Obta
inin
g T
h
ro
ug
h Reusability
in Pro
d
uct Line
Th
e t
o
tal b
e
n
e
fit o
f
reusab
ility is ob
tain
ed
thro
ugh
th
e fo
llowing
eq
u
a
tion
:
B
∑
|
|
∑
|
|
B
(1
0)
Accord
ing
to
th
is eq
u
a
tion
t
h
e to
tal b
e
n
e
fi
t o
b
t
ain
i
ng
thro
ugh
th
e reu
s
i
n
g
o
f
assets i
n
software
p
r
od
u
c
t lin
e is eq
u
a
l w
ith
t
h
e b
e
n
e
f
it o
b
t
ai
n
i
ng
th
rou
g
h
t
h
e r
e
u
s
ing
of
assets in
all in
d
i
v
i
du
al pr
oducts o
f
pr
o
duct
l
i
n
e
.
3
.
4
.
2
.
T
h
e Impa
ct
o
f
the Reusa
b
ility
o
f
an
Asset o
n
Develo
pi
ng a Giv
e
n Pro
duct
Th
e im
p
act o
f
th
e reusab
ility o
f
th
e asset aj
o
n
t
h
e d
e
v
e
lopin
g
o
f
th
e
p
r
odu
ct p
i
is d
e
ri
v
e
d
th
rou
g
h
th
e fo
llowing
eq
u
a
tion
:
I
∗
(1
1)
In
wh
ich
Iij
is th
e i
m
p
act o
f
th
e reu
s
ab
ility
o
f
th
e asset aj
in
d
e
v
e
l
o
p
i
n
g
t
h
e produ
ct p
i
, w(aj
) is th
e
wei
g
ht
val
u
e
o
f
t
h
e asset
aj,
ki
j i
s
t
h
e pe
rce
n
t
of c
h
a
nges
of t
h
e ass
e
t
aj
appl
y
i
n
g
f
o
r
u
s
i
ng i
n
t
h
e p
r
o
duct
pi
an
d
w(p
i
) is th
e weig
h
t
v
a
l
u
e o
f
to
tal assets u
s
ed
in
th
e pr
o
duct
pi
. I
f
we
wi
sh t
o
ex
p
r
es
s t
h
e wei
ght
va
l
u
e of
assets in
term
s o
f
lin
e
o
f
cod
e
, we ca
n
rewrite th
e eq
u
a
ti
o
n
(1
1) as fo
llows:
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
Metric su
ite to
Eva
l
ua
te Reu
s
a
b
ility o
f
So
ftwa
r
e Produ
ct Li
n
e
(Moha
mmad
Ali To
rka
m
an
i)
28
9
I
∗
(1
2)
In wh
ich Ii
j
is
th
e im
p
act o
f
t
h
e
reu
s
ab
ility o
f
th
e asset aj
in
d
e
v
e
l
o
p
i
n
g
th
e pro
d
u
c
t
p
i
, size (aj
)
is
th
e nu
m
b
er
o
f
th
e lin
es
of th
e asset aj
,
k
ij is th
e
p
e
rc
e
n
t
of
cha
n
g
e
s
of
t
h
e asset
a
j
a
ppl
y
i
ng
fo
r
usi
n
g
i
n
t
h
e
p
r
od
u
c
t and
size(p
i) is th
e num
b
e
r o
f
t
h
e lines of th
e pro
duct p
i
.
Exam
ple: s
u
ppos
e t
h
at
size
p
2
K
L
O
C
و
size
a
400LOC
،
k
11
20%
.
From
th
e equ
a
tion
(1
1)
we
ha
ve:
I
400
∗
1
20
100
200
0
0
.
1
6
Al
so,
we
ca
n c
a
l
c
ul
at
e I1
1 t
h
r
o
u
g
h
t
h
e
eq
uat
i
on
(
1
2
)
as
f
o
l
l
o
ws:
I
400
400
∗
0
.2
20
00
0
.
1
6
3
.
4
.
3
.
T
h
e Impa
ct
o
f
Reusability
o
n
Dev
e
lo
ping a Giv
e
n Pro
duct
Th
e im
p
act o
f
th
e reusab
ility o
f
assets
o
n
d
e
v
e
lop
i
ng
th
e
p
r
o
d
u
c
t p
i
is d
e
riv
e
d
t
h
ro
ugh
the fo
llowing
equat
i
o
n:
I
∑
|
|
I
(1
3)
In wh
ich Ii is t
h
e im
p
act o
f
reu
s
ab
ility o
n
the produ
ct
p
.
3
.
4
.
4
.
T
h
e
Impa
ct o
f
Reusability
o
n
Dev
e
lo
ping All Products
of Pro
d
uct Line
Th
e im
p
act o
f
reu
s
ab
ility o
n
d
e
v
e
l
o
p
i
n
g
all
p
r
od
u
c
ts
o
f
so
ftware pro
d
u
c
t
lin
e is calcu
lated
th
rou
g
h
th
e fo
llowing
eq
u
a
tion
:
∑
|
|
(1
4)
In
wh
ich
Ii is
th
e im
p
act o
f
reu
s
ab
ility o
n
th
e pro
d
u
c
t
p
i
.
W
e
can
rewri
t
e th
e abov
e eq
u
a
tion
as
fo
llows:
I
∑∑
|
|
|
|
I
(1
5)
In
wh
ich
Ii
j
is th
e i
m
p
act o
f
th
e reu
s
ab
ility o
f
th
e asset
aj
on
th
e
d
e
v
e
lo
p
i
ng
o
f
th
e
p
r
od
u
c
t p
i
.
Im
p
act o
f
Reusab
ility
m
easu
r
es
reu
s
e b
e
n
e
fit o
f
so
ftwa
re produ
ct lin
e.
Norm
all
y
so
ft
ware produ
ct l
i
n
e
h
a
s
m
o
re
m
e
m
b
ers, th
is m
e
tric is b
i
gg
er
, an
d prod
u
c
t lin
e is m
o
re eco
n
o
m
ic.
4.
CASE ST
UDY
In th
is secti
o
n
,
Ou
r Metric su
ite is ev
al
uated in practice in
Ira
nian Telecom
m
unication
Man
u
f
actur
ing Co
m
p
an
y (I
TMC)
. ITMC
is
a co
m
p
any operating in Electri
cal engine
ering and ICT
areas.
Beside som
e
products in electrical a
nd com
m
uni
cat
i
on area, ITM
C
i
s
devel
o
pi
n
g
som
e
soft
wa
re sy
st
em
s. In
or
der
t
o
t
a
ke a
dva
nt
age
of
S
o
ft
ware
Pr
o
d
u
c
t
Li
ne, R
&
D
depa
rt
m
e
nt
of ITM
C
ha
s dev
e
l
ope
d fi
ve
So
ft
wa
r
e
pr
o
duct
l
i
n
es:
SPL1:
S
o
ft
wa
r
e
Pr
o
duct
Li
ne
fo
r M
o
bi
l
e
Set
s
SPL2:
softwa
re product line
for Telecomm
unication Ce
nters
SPL3:
S
o
ft
wa
r
e
Pr
o
duct
l
i
n
e
f
o
r
EC
U (
an
d
Sm
art
cont
rol
s
y
st
em
s for ca
rs
)
SPL4:
S
o
ft
wa
r
e
Pr
o
duct
l
i
n
e
f
o
r
ATM
an
d B
a
nki
ng
sy
st
em
s
SPL5:
ERP
Software
P
r
oduct line
Th
e ev
al
u
a
tio
n in
d
e
x
e
s em
p
l
o
y
in
g
in
th
is case stu
d
y
are SCC (Stru
c
t
u
ral Si
milarity
Co
efficien
t),
RBR (Reu
se B
e
n
e
fit Rate) PrR (Prod
u
c
t
-rel
a
ted
Reu
s
ab
ility
) and
SOC (Size o
f
C
o
mm
o
n
ality). Tab
l
es
2
to
7
sho
w
t
h
e
dat
a
bel
o
ngi
ng t
o
t
h
e p
r
o
d
u
ct
l
i
n
e
1. Ta
bl
e 2 s
h
ows t
h
e l
i
s
t
of
t
h
e com
m
on asset
s
of t
h
e S
o
ft
ware
pr
o
duct
l
i
n
e ar
chi
t
ect
ure 1
.
T
a
bl
e 3 sh
o
w
s t
h
e l
i
s
t
of asset
s
reuse
d
i
n
s
o
m
e
pro
duct
s
of
t
h
e soft
ware
p
r
o
d
u
ct
l
i
n
e 1. Tabl
e 4
sho
w
s t
h
e l
i
s
t
of ot
her
new
-
devel
ope
d asse
t
s
. Tabl
e 5 sh
ows t
h
e l
i
s
t
of t
h
e pr
od
uct
s
of t
h
e
so
f
t
w
a
r
e
pr
odu
ct lin
e
1
along
w
ith
th
e info
r
m
atio
n
of
ever
y pr
odu
ct inclu
d
i
ng
pr
oduct n
a
m
e
, th
e na
m
e
o
f
assets reuse
d
in products, the weight
value
of each a
sset, asset type (developed
or
reused) a
nd
perce
n
t
of
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
IJEC
E V
o
l
.
4, No
. 2, A
p
ri
l
20
14
:
28
5 – 2
9
4
29
0
changes a
ppli
e
d for adopting with ne
w a
r
chitecture an
d
pr
odu
ct w
e
ig
h
t
(W
p)
.
A
l
so
in
th
ese table, th
e
calcu
latio
n
s
of m
e
trics
W
t
%Rp
an
d
Ao
RP, th
e im
p
act o
f
th
e reu
s
ab
ility
o
f
an
asset on
p
r
od
u
c
t
d
e
v
e
lop
m
en
t
(Iij), th
e im
p
act o
f
reu
s
ab
ility o
n
d
e
v
e
l
o
p
i
ng
a g
i
v
e
n
p
r
o
d
u
c
t (Ii), th
e im
p
act o
f
reu
s
ab
ility o
n
all p
r
o
d
u
c
ts an
d
p
r
od
u
c
t related reu
s
ab
ility(PrR). At first we calcu
lated
ev
alu
a
tio
n
ind
e
x
e
s an
d
th
e m
e
tri
c
s in
trod
u
c
ed
in
th
is
pape
r. T
h
e
obt
ai
ned r
e
sul
t
s
w
e
re save
d i
n
t
a
bl
e 6. Ta
bl
e 7
defi
nes t
h
e ra
n
k
s o
f
p
r
od
uct
l
i
n
es 1 t
o
5 f
o
r
eac
h
com
p
ari
s
on
as
pect
. T
h
i
s
t
a
bl
e has
bee
n
pre
p
ared
usi
n
g
t
h
e
resul
t
s
sa
ve
d i
n
t
h
e
t
a
bl
e
6.
Tabl
e
2. T
h
e
l
i
s
t
of
com
m
on
asset
s
o
f
the
architecture
of
the pr
odu
ct lin
e
1
no
asset
wi
ty
pe
Per
cept of Changes
1
a1
1
Reuse
20
2
a2 2
Reuse
BB
3
a3
2
Reuse
BB
4
a4 1
Reuse
BB
5
a5
7
Reuse
40
6
a6 3
Reuse
40
7
a7
5
Develop.
8
a8 2
Develop.
Tabl
e 3.
Th
e list o
f
assets
reused
i
n
so
m
e
p
r
o
d
u
c
ts
o
f
th
e
produ
ct lin
e
1
no asset
w
i
1
a9
2
2
a10 1
3
a11
1
4
a12 2
5
a13
1
6
a14 1
7
a15
1
8
a16 2
9
a17
2
10
a18 2
11
a19
2
12
a20 2
Tabl
e
4.
Th
e l
i
s
t
of
ne
w
de
vel
ope
d a
sset
s
of t
h
e
pr
o
duct
l
i
n
e
1
n
o
as
s
e
t
wi
r
o
w
A
s
s
e
t
wi
r
o
w
as
s
e
t
wi
r
o
w
as
s
e
t
wi
1
a21
1
11
a31
1
21
a41
1
31
a51
1
2
a22 1
12
a32
1
22
a42
1
32
a52
1
3
a23
1
13
a33
1
23
a43
1
33
a53
1
4
a24 1
14
a34
1
24
a44
1
34
a54
1
5
a25
1
15
a35
1
25
a45
1
35
a55
1
6
a26 1
16
a36
1
26
a46
1
36
a56
1
7
a27
1
17
a37
2
27
a47
1
37
a57
1
8
a28 1
18
a38
5
28
a48
1
38
a58
1
9
a29
3
19
a39
1
29
a49
1
39
a59
1
10
a30 1
20
a40
1
30
a50
1
40
a60
1
If y
o
u
com
p
ar
e ro
ws 9 an
d
13 i
n
t
h
e t
a
b
l
e 7, y
ou wi
l
l
fi
nd t
h
at
t
h
e
m
e
t
r
i
c
of “t
he im
pact
of
reusa
b
ility on
products” is c
o
m
p
letely
in accorda
n
ce wit
h
the m
e
tric of RBR. Our m
e
tric has a
n
advantage
co
m
p
ared
wit
h
th
e m
e
tric
o
f
RBR. It can
b
e
calcu
lated
fo
r each
pro
d
u
c
t and
is n
o
t
gen
e
ral lik
e RBR.
Co
m
p
ariso
n
of th
e rows 10 an
d
14
o
f
the tab
l
e 7
rev
e
als th
at th
e re
su
lts o
f
th
e metric o
f
W
t
%Rp
a
re
co
m
p
letely si
milar to
th
e averag
e ob
tain
ed th
rou
g
h
Pr
R
m
e
t
r
i
c
. Al
so
o
u
r
W
t
%R
p m
e
t
r
i
c
gi
ve
s t
w
o
di
ffe
re
nt
v
a
lu
es to
the produ
ct lin
es 2
an
d
3
wh
ile th
eir v
a
lu
es
in
PrR
m
e
tric are t
h
e sam
e
. Th
e resu
lts of th
is
case
stu
d
y
show that th
e p
r
odu
ct lin
e 2
h
a
s the
m
a
x
i
m
u
m
weigh
t
v
a
lu
e i
n
reu
s
ab
ility as it h
a
s g
a
in
ed
th
e
max
i
m
u
m
v
a
l
u
e in
W
t
%R
pmetric. In
o
t
her wo
rd
s, in
t
h
is p
r
od
uct li
n
e
th
e ratio
of th
e wei
g
h
t
s
o
f
the
com
pone
nt
s o
f
com
m
on part
of s
o
ft
war
e
p
r
od
uct
l
i
n
e a
r
ch
i
t
ect
ure t
o
t
h
e
wei
g
ht
o
f
wh
o
l
e com
pone
nt
s of t
h
e
so
ft
ware
p
r
o
duct lin
e g
a
in
s th
e
m
a
x
i
m
u
m
v
a
lu
e. For th
is reason
, it is ex
p
ected
th
at th
e p
r
o
d
u
c
t lin
e 2
wi
ll b
e
m
o
re su
ccessfu
l
fro
m
th
e v
i
ewpo
in
t of reusab
ility. (Fo
r
ex
am
p
l
e co
m
p
are weigh
t
v
a
lue with
th
e n
u
m
b
er o
f
l
i
n
es o
f
t
h
e
p
r
o
g
ram
or
re
qui
r
e
d e
f
f
o
rt
fo
r
de
vel
o
pi
n
g
pu
r
p
o
s
es.)
In
or
de
r t
o
c
h
eck t
h
e acc
u
r
a
c
y
of t
h
e m
e
t
r
i
c
of
AoR
P
, c
o
m
p
are t
w
o
di
ffe
rent
pr
o
duct
s
wi
t
h
eac
h
ot
he
r (
f
o
r e
x
a
m
pl
e pro
duct
s
p1
an
d
p
2
bel
o
ngi
ng
t
o
t
h
e
pr
od
uct
l
i
n
e
1
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
Metric su
ite to
Eva
l
ua
te Reu
s
a
b
ility o
f
So
ftwa
r
e Produ
ct Li
n
e
(Moha
mmad
Ali To
rka
m
an
i)
29
1
As you
m
a
y se
e in
th
e tab
l
e 5, th
e m
o
st co
mp
on
en
ts of th
e
p
r
od
u
c
t
p
1
are
Black
Bo
x
type. Gen
e
rally
th
e co
m
p
on
en
t
s
of
th
e pr
odu
ct p
1
r
e
qu
ir
e
f
e
w
e
r
ch
ang
e
s
f
o
r
ad
op
ting
w
ith th
e ar
ch
itecture of
th
e pr
odu
ct lin
e
com
p
ared
with the produ
ct p2
. Th
e m
e
tric
o
f
Ao
RP sh
ows th
is
fact
clearly. (Th
e
v
a
lu
e of th
is m
e
tric is
0
.
8
545
45
fo
r
t
h
e p
r
od
u
c
t p1
an
d
0.745
445
fo
r
th
e p
r
od
u
c
t p
2
.)
Thi
s
m
e
t
r
i
c
wi
ll
wor
k
f
o
r ot
h
e
r com
pone
nt
s t
oo e
v
e
n
i
f
y
o
u
sel
ect
i
t
e
m
s
fro
m
di
ffere
nt
p
r
o
duct
l
i
n
es
.
Th
e
m
e
tric
o
f
Ao
RSPL is similar
to
Ao
RP. Th
e o
n
l
y d
i
ff
eren
ce is th
at th
is
m
e
tric esti
mates o
n
l
y th
e co
mm
o
n
com
pone
nt
s of
t
h
e archi
t
ect
u
r
e of
pr
od
uct
l
i
n
e. Acc
o
r
d
i
n
g
t
o
t
h
e t
a
bl
e 7, am
ong va
ri
o
u
s
pr
od
uct
l
i
n
es, t
h
e
pr
o
duct
l
i
n
e 2 has t
h
e m
a
ximum
AoR
SPL. Thi
s
m
eans t
h
at
t
h
e co
m
m
on com
ponent
s o
f
t
h
e archi
t
ect
ure
o
f
t
h
e pr
o
d
u
c
t
l
i
n
e 2 re
qui
re fe
wer c
h
an
ges
f
o
r a
d
o
p
t
i
n
g wi
t
h
ne
w arc
h
i
t
ect
ure c
o
m
p
ared wi
t
h
ot
he
r p
r
o
d
u
ct
lin
es.
Tabl
e 5.
Th
e pr
odu
cts of
t
h
e
p
r
od
u
c
t lin
e 1
Pr
oduct
asset
wi Type
Percent
of
Changes
1
100
Wp
Wt%
R
p
AoR
P
I
ij
I
i
I
PrR
R:
Reuse
D:Deve
lp
m
e
nt
p
1
a1 1
R
20
0.
8
29
22
75.
862
068
97
0.
8545
454
55
0.
0275
862
07
0.
5931
034
48
2.
1432
000
67
0.
6153
846
15
a2 2
R
BB
1
0.
0689
655
17
a3 2
R
BB
1
0.
0689
655
17
a4 1
R
BB
1
0.
0344
827
59
a5 7
R
40
0.
6
0.
1448
275
86
a6 3
R
40
0.
6
0.
0620
689
66
a7 5
D
a8 2
D
a11 1
R
20
0.
8
0.
0275
862
07
a12 2
R
BB
1
0.
0689
655
17
a13 1
R
BB
1
0.
0344
827
59
a14 1
R
BB
1
0.
0344
827
59
a15 1
R
40
0.
6
0.
0206
896
55
p
2
a1 1
R
20
0.
8
45
38
84.
444
444
44
0.
7454
545
45
0.
0177
777
78
0.
4044
444
44
0.
3478
260
87
a2 2
R
BB
1
0.
0444
444
44
a3 2
R
BB
1
0.
0444
444
44
a4 1
R
BB
1
0.
0222
222
22
a5 7
R
40
0.
6
0.
0933
333
33
a6 3
R
40
0.
6
0.
04
a7 5
D
a8 2
D
a16 2
R
10
0.
9
0.
04
a17 2
R
60
0.
4
0.
0177
777
78
a18 2
R
25
0.
75
0.
0333
333
33
a19 2
R
40
0.
6
0.
0266
666
67
a20 2
R
45
0.
55
0.
0244
444
44
a21 1
D
a22 1
D
a23 1
D
a24 1
D
a25 1
D
a26 1
D
a27 1
D
a28 1
D
a29 3
D
a30 1
D
p
3
a1 1
R
20
0.
8
30
18
60
0.
95
0.
0266
666
67
0.
45
0.
5714
285
71
a2 2
R
BB
1
0.
0666
666
67
a3 2
R
BB
1
0.
0666
666
67
a4 1
R
BB
1
0.
0333
333
33
a5 7
R
40
0.
6
0.
14
a6 3
R
40
0.
6
0.
06
a7 5
D
a8 2
D
a9 2
R
15
0.
85
0.
0566
666
67
a31 1
D
a32 1
D
a33 1
D
a34 1
D
a35 1
D
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
IJEC
E V
o
l
.
4, No
. 2, A
p
ri
l
20
14
:
28
5 – 2
9
4
29
2
p
4
a1 1
R
20
0.
8
38
21
55.
263
157
89
0.
78
0.
0210
526
32
0.
4
0.
4705
882
35
a2 2
R
BB
1
0.
0526
315
79
a3 2
R
BB
1
0.
0526
315
79
a4 1
R
BB
1
0.
0263
157
89
a5 7
R
40
0.
6
0.
1105
263
16
a6 3
R
40
0.
6
0.
0473
684
21
a7 5
D
a8 2
D
a9 2
R
40
0.
6
0.
0315
789
47
a10 1
R
40
0.
6
0.
0157
894
74
a11 1
R
30
0.
7
0.
0184
210
53
a12 1
R
10
0.
9
0.
0236
842
11
a36 1
D
a37 2
D
a38 5
D
a39 1
D
a40 1
D
p
5
a1 1
R
20
0.
8
46
19
41.
304
347
83
0.
775
0.
0173
913
04
0.
2956
521
74
0.
2666
666
67
a2 2
R
BB
1
0.
0434
782
61
a3 2
R
BB
1
0.
0434
782
61
a4 1
R
BB
1
0.
0217
391
3
a5 7
R
40
0.
6
0.
0913
043
48
a6 3
R
40
0.
6
0.
0391
304
35
a7 5
D
a8 2
D
a10 1
R
40
0.
6
0.
0130
434
78
a17 2
R
40
0.
6
0.
0260
869
57
a41 1
D
a42 1
D
a43 1
D
a44 1
D
a45 1
D
a46 1
D
a47 1
D
a48 1
D
a49 1
D
a50 1
D
a51 1
D
a52 1
D
a53 1
D
a54 1
D
a55 1
D
a56 1
D
a57 1
D
a58 1
D
a59 1
D
a60 1
D
Tabl
e
6.
C
a
l
c
u
l
at
i
on o
f
m
e
t
r
i
c
s an
d c
o
m
p
ari
s
on
of t
h
e
resu
l
t
s ob
tain
ed fo
r
p
r
od
u
c
t lin
e 1 to
5
no
Co
m
p
ar
ison
aspect
SPL
1
SPL
2 SPL
3 SPL
4
SPL
5
1
T
o
tal nu
m
b
er
of co
m
ponents
60
40
40
44
16
2
T
h
e nu
m
b
er
of com
ponents used in
co
m
m
on
architecture s
ection
6
10
10
15
5
3
nu
m
b
er of pr
oduct
s
5
6
6
10
5
4
SOC
8
10
10
15
5
5 SSC
0.
1333
333
33
0.
25
0.
25
0.
3409
09
0.
3125
6 W
s
pl
89
72
96
105
44
7
T
h
e weight of the m
e
m
b
er
s of pr
oduct line
188
206
233
725
104
8
T
h
e weight of the co
m
ponents of com
m
o
n
Par
t
of
architecture
23
31
29
61
13
9 RBR
2.1123
595
51
2.8611
11
2.
4270
83
6.9047
62
2.3636
36
10
T
h
e aver
age of PrR
0.
4543
788
35
0.
6570
2
0.
6570
2
0.
7275
46
0.
6607
14
11
The average of
Ao
R
P
0.
821
0.
7064
01
0.
8416
57
0.
7447
31
0.
6583
33
12
AoR
SP
L
0.
625
0.
875
0.
725
0.
7166
67
0.
66
13
I
2.
1432
000
67
3.
7882
22
3.
5663
79
7.
0078
18
2.
9413
45
14
The average of the
weight
percent of
reusability
of pr
od
ucts (
W
t%
Rp)
63.
374
803
83
82.
679
09
76.
147
7
96.
503
85
83.
955
62
15
W
t
%R
12.
234
042
55
15.
048
54
12.
446
35
8.
4137
93
12.
5
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
Metric su
ite to
Eva
l
ua
te Reu
s
a
b
ility o
f
So
ftwa
r
e Produ
ct Li
n
e
(Moha
mmad
Ali To
rka
m
an
i)
29
3
Table 7
.
The ranks of diff
erent
aspects of
th
e pr
oduct
lines 1
to
5
No Co
m
p
arison
aspe
ct
Rank
1
Rank
2
Rank
3
Rank
4
Rank
5
1
T
o
tal nu
m
b
er
of co
m
ponents
1
4
2,
3
5
2
T
h
e nu
m
b
er
of com
ponents used in
co
m
m
on
architecture s
ection
4 2,
3
1
5
3
nu
m
b
er of pr
oduct
s
4
2,
3
1,
5
4
SOC 4
2,
3
1
5
5
SSC
4
5
2,
3
1
6
W
s
pl 4
3
1
2
5
7
T
h
e weight of the m
e
m
b
er
s of pr
oduct line
4
3
2
1
5
8
T
h
e weight of the co
m
ponents of com
m
o
n
Part of
architectur
e
4 2
3
1
5
9
RBR
4
2
3
5
1
10
T
h
e aver
age of PrR
4
5
2,
3
1
11
The average of
Ao
R
P
3
1
4
2
5
12
AoR
SP
L
2
3
4
5
1
13
I
4
2
3
5
1
14
The average of
the
weight percent of
reusability
of products (W
t%Rp)
4 5
2
3
1
15
Wt%
R
2
5
3
1
4
5.
CO
NCL
USI
O
N
We argu
ed
t
h
at th
e m
o
st of av
ailab
l
e m
e
tric
s em
pl
oy
i
ng
fo
r est
i
m
ati
ng
pr
od
uct
l
i
n
e
arc
h
i
t
ect
ure ar
e
insufficient a
n
d also em
ploying thes
e m
e
tr
ics are diffic
ul
t. Product line
architectur
e i
s
a field with
fewe
r
m
e
t
r
i
c
s. In
re
cent
y
ears
s
o
m
e
new
m
e
t
r
ics ha
ve
bee
n
pr
o
pose
d
f
o
r
est
i
m
a
ti
ng
pr
o
duct
l
i
n
e a
r
chi
t
ect
ure.
Al
t
h
o
u
gh t
h
e p
r
o
p
o
sed m
e
t
r
i
c
s are use
f
ul
t
h
ey
have
not
be
en wi
del
y
used
i
n
i
n
d
u
st
ri
es.
For t
h
i
s
, e
xpe
r
t
s and
R
&
D
de
part
m
e
nt
s s
h
o
u
l
d
pay
m
o
re at
t
e
nt
i
on
t
o
t
h
e m
e
t
r
i
c
s em
pl
oy
i
ng i
n
pr
od
uct
l
i
n
e
arc
h
i
t
ect
ure.
In
so
ft
wa
re
pr
od
uct
l
i
n
e
,
var
i
ous
t
y
pes
of
asset
s
are
bei
n
g
u
s
ed
. T
h
e
val
u
es
o
f
t
h
es
e asset
s
a
r
e
d
i
fferen
t
and
also
th
e profit ob
tain
ing
thro
ugh
u
s
ing
th
ese assets is
differen
t
.
Desp
ite o
f
t
h
is
fact, m
o
st
av
ailab
l
e m
e
tr
i
c
s don
’
t
co
nsider
th
e w
e
i
g
h
t
valu
es
o
f
t
h
e assets of
so
f
t
w
a
re pr
odu
ct lin
e.
W
e
p
r
op
o
s
ed
i
n
our
p
a
p
e
r so
m
e
n
e
w m
e
trics fo
r
esti
m
a
t
i
n
g
reu
s
ab
ility in
so
ft
ware pro
d
u
c
t lin
e. Th
ese m
e
trics
co
nsid
er th
e
weig
h
t
values
of asset
s
.
Our Metric su
i
t
e is ev
alu
a
ted in
p
r
actice in
Ira
ni
an Tel
e
c
o
m
m
uni
cat
i
on M
a
nu
fact
uri
n
g C
o
m
p
any
.
Alon
g with o
t
her m
e
trics, o
u
r
p
r
op
o
s
ed
m
e
trics can h
e
l
p
u
s
t
o
estim
ate th
e q
u
a
lity of so
ft
ware produ
ct lin
e.
REFERE
NC
ES
[1]
Bas
s
L,
Clem
ent
s
P
and Ka
zm
an
R.
Sof
tware Arc
h
ite
cture in Prac
tic
e
. Second
Edition, Addison Wesley
.
2003
.
[2]
van der Hoek, A Dincel E and Medvidoviü N. “
Using Service Utilization Metricst
o Assess the Stru
cture of Product
L
i
ne Ar
chit
ectur
es
”. Proceedings
of the Ninth In
ternational So
ftware Metrics S
y
mposium (METRICS’03), IEEE.
2003.
[3]
Zhang T
,
Deng
L, W
u
J, Zhou
Q and
Ma C. “
S
om
e Metrics for
Accessing Qual
it
y
of Product L
i
neArchi
t
ec
ture
”
.
IEEE
. Vol 27, Issue 3, pp 35
– 4
1
, 2008
.
[4]
Berger C And Busse C. “Produc
t Line
Metrics f
o
r Legacy
Softw
a
re in Prac
tice”. The 14th Intern
ation
a
l Software
Product Line. 20
10.
[5]
Berger C,
Rend
el H
and
Rumpe B. “
Measuring the Ability to
Form a
Product Line from Existing Products
”.
Variabi
lit
y
Mod
e
lling
of Softw
a
r
e
-int
ensive S
y
stem
s (VaMos). 2010.
[6]
SEI, A Framework for Software
Product Lin
e
Practice,
Version 5.
0, 2009, Available:
http://www.sei.cmu.e
du/productlines/frame_rep
o
r
t/miningEAs.htm
[7]
Aldekoa G,
Tr
ujillo
S, Sag
a
r
dui G and
D
í
az
O.
EXPERIENCE MEASUR
ING MAINTAINABILITYIN
SOFTWARE PRODUCT LINES.
XV Jornadas
de Ingen
i
ería
del So
ftware y Bas
e
s de Datos
(
J
ISBD)
. 2006.
[8]
Rahman A. Metrics for the Structural Assessment of
Prod
uct Lin
e
Architectur
e. MSC Thesis, School of
Engineering
at B
l
eking
e
Institu
te of Techno
log
y
,
Sweden. 2004
.
[9]
Junior E, Gimen
e
s I and Mal
don
ado J. “A Metric Suite to Support Softwa
re Product Lin
e
Archit
ecture Evaluation”.
XXXI
V Conf
eren
ciaLatinoameric
ana de In
formáti
c
a (
C
LEI 2008)
. 2008.
[10]
Zubrow D and Chastek G. Measures fo
r Software Product Lines
.
Technical
R
e
p
o
rt, Carn
egie M
e
llon University
.
2003.
[11]
Benavid
e
s D,
T
r
inidad
P and
R
u
iz-cor
tés A. “
A
utomated Reasoning onFeatu
r
e
Models”.
LNCS, Advanced
Information
S
y
s
t
ems Engin
eer
in
g: 17
thInternational Conf
er
en
ce, CAISE. 2005.
[12]
Mendonca M,
Wasowski A, Czarnecki K
an
d Cowan D. “
Effi
ci
ent
compil
ation
te
chnique
s for
large sc
al
e
feature models
”. In
Gen
e
rativ
e Programming and
Compone
nt Engineer
ing,
7th International
Conference,GPCE , Procee
dings.
pages 13–22
, 20
08.
[13]
Fernandez-Amoros D, Gil R
and So
moli
nos J.
“Inferring Info
rmation fro
m Fe
ature Diagr
a
ms
to Product Line
Econom
ic Models”. ACM Internat
ional Conf
erenc
e
Pr
oceed
i
ng Series Vol. 446, Proceedi
ngs of the 13th
Internationsl Sof
t
ware Product
Li
ne Confer
ence, p
p
. 41-50
, 2009
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
IJEC
E V
o
l
.
4, No
. 2, A
p
ri
l
20
14
:
28
5 – 2
9
4
29
4
[14]
Czarne
cki K and Kim
P
.
“
C
ardinalit
y-b
a
s
e
d feat
ure m
ode
ling and cons
traints
:
A progres
s
report”. In P
r
oceed
in
gs
of the Internatio
nal Workshop o
n
Soft
ware Factories At OOPSLA 2005. 2005
.
[15]
Kas
i
kci B and B
ilgen S
.
“
S
cal
ab
le m
ode
ling of
software prod
uct lin
e va
r
i
ability
”. 13th Intern
ation
a
l Software
Product Line Conference, SPLC
2009,
SanFran
c
isco. August 200
9.
[16]
Benavid
e
s D, Segura S and Ruiz-Cortés A.
“
A
utom
ated anal
ysis of feature
m
odels 20 y
e
ars later
:
A litera
tu
re
review”.
Information Sy
ste
m
s.
2
010.
[17]
Jones LG. Prod
uct
Line Acquisition in
th
e DoD: The Prom
is
e, T
h
e Cha
lleng
es
,
Techn
i
ca
l Not
e
,
Carnegi
e
M
e
l
l
on
University
,1999
.
BI
O
G
R
A
P
HY
OF
A
U
T
HO
R
Mohammad Ali
Torkamaniwas b
o
rn in Iran, Shir
az
City
, in 1975
. He receiv
ed th
e M.S. degr
ee
in software engineering from th
e Shahid Behes
h
ti University
, in 2011. He is the author of 15
books (in Persian), more than 35 ar
ticles. His research in
terests
include softwar
e
architecture,
Ultra
Larg
e S
c
al
e s
y
s
t
em
s
,
cr
yp
t
ograph
y
an
d
Network security
an
d holds one patent.
He is working
in R&D Department of Ira
nian
Telecommunication Manufactu
ring Compan
y
now. Also, he is currently
teaching at the Univ
ersity
of Applied Science and
Techno
log
y
in
S
h
iraz.
Evaluation Warning : The document was created with Spire.PDF for Python.