=Paper=
{{Paper
|id=Vol-2207/IWSM_Mensura_2018_paper_3
|storemode=property
|title=Automated Size Measurement of Embedded System based on XML using COSMIC FSM
|pdfUrl=https://ceur-ws.org/Vol-2207/IWSM_Mensura_2018_paper_3.pdf
|volume=Vol-2207
|authors=Thandar Zaw,Swe Zin Hlaing,Myint Myint Lwin,Koichiro Ochimizu
|dblpUrl=https://dblp.org/rec/conf/iwsm/ZawHLO18
}}
==Automated Size Measurement of Embedded System based on XML using COSMIC FSM==
https://ceur-ws.org/Vol-2207/IWSM_Mensura_2018_paper_3.pdf
Automated Size Measurement of Embedded System
based on XML using COSMIC FSM
Thandar Zaw, Swe Zin Hlaing, Myint Myint Lwin, and Koichiro Ochimizu
University of Information Technology, Myanmar
thandarzaw@uit.edu.mm,swezin@uit.edu.mm,
myintmyintlwin@uit.edu.mm, ochimizu@jaist.ac.jp
Abstract. Nowadays, function size is an important role for measuring software
before the development phase in software development life cycle. Industries
need well-defined different notations and measurement method. Thus,
COSMIC FSM is one of the International Standards of FSM for measuring the
size of embedded software. Manual measurement is time consuming and com-
plicated to measure the size of large system. This paper proposes the automated
measurement tool that accepts the XML documents from three different nota-
tions such as the UML, SysML and Petri net. Then, the XML documents of
these diagrams have been translated to COSMIC FSM by using mapping rules
with the case study of cooker system. This paper shows the result of software
size by using the prototype tool that is the same with the result of manual meas-
urement.
Keywords Common Software Measurement International Consortium
(COSMIC FSM), Sequence diagram, UML, SysML, Petri net
1 Introduction
Software size estimation is an important input for estimating the effort required to
develop the measured software. Software sizing is the prediction of the software size
to build the system. Several size measurement methods have been proposed to meas-
ure the size of software. The most popular size estimation methods are source line of
code (SLOC) and functional size measurement (FSM). Size-related measure is based
on some output from the software process and it is language dependence. And func-
tion-related measure such as functional measurement (FSM) is based on an estimate
of the functionality of the delivered software. Function size measurement methods are
technology independent and can be estimated early in analysis and design.
Functional Size Measurement (FSM) is a great method in industrial development
as it supports the input to estimate the effort. Software functional size measurement
methods that have been recognized as various ISO FSM: IFPUG FPA, MKII,
NESMA, FISMA, and COSMIC FSM[1]. The other traditional methods are applied
only in business software but they are hardly applied in real-time software and em-
bedded software. But COSMIC can be applied in both software.
14
T. Zaw, S. Z. Hlaing, M. M. Lwin, K. Ochimizu
Many researchers proposed the software estimation methods which are not appli-
cable for various modeling notations. To address this limitation, this paper proposed
the automated software size measurement tool for different modeling notations to
estimate the functional size of software. These notations have been translated to
COSMIC by using mapping rules with a simple case study of cooker system. This
paper is organized as follows: the second section provides related work; the third
section presents the proposed system; the fourth section explains the evaluation and
final section is conclusion and future work.
2 Related Work
In [2], Symons, C. described the COSMIC concepts that can be applied in any real-
time software requirements to measure the functional size of real-time software to
understand clearly for any software engineer with alarm example. In [3], Soubra, H.,
et al. proposed the design of the FSM procedure based on the documentation of the
mapping of the Simulink concepts to COSMIC concepts for the embedded real-time
software system. In [5], Luigi Lavazza., et al proposed the UML that can be used to
build models according to the COSMIC measurement rules. Asma Sellami et al. [6]
proposed the measurement method for sizing of sequence diagram that can be meas-
ured both the functional and structural size at different level of granularity. In [4], the
author proposed the automated functional and structural measurement of software size
from XML structure of sequence diagram to calculate COSMIC CFP. This paper
proposed the automated FSM tool which is based on the XML documents from dif-
ferent design model notations. Then, the mapping rules are defined between these
notations and COSMIC. It is expected that the result of the functional size of software
can be provided by many industries to increase effort and productivity.
3 Proposed System
This section describes the proposed method to calculate the functional size of soft-
ware by using the XML structure of three different notations. The proposed system of
three phases is shown in Fig 1. In measurement strategy, the proposed system is ana-
lyzed XML document from the popular notations such as UML and SysML sequence
diagram and Petri net. In mapping phase, after analyzing the design models with
COSMIC concepts, the design models have been translated into COSMIC by using
the mapping rules. In measurement phase, the actual size of software is calculated by
using COSMIC method.
15
IWSM/Mensura’18, September 18–20, 2018, Beijing, China
Fig. 1. Proposed System based on different design notations
3.1 UML Use Case Model of Cooker System as Functional User Requirements
This paper discusses the specification of a simple version of the cooker system which
is used as a case study to determine the counting of COSMIC [6]. Before developing
the UML, SysML and Petri net representation, the specification of the cooker system
must be defined. The functional user requirements of this system are as follows:
1. The cooker software can get the input from a door sensor and start button. Then, it
can show the light and heater on/off when the power is switched on.
2. When the start button is pressed and the door is closed, the cooking starts. If the
door is open, the start button has no effect.
3. Either the door is open while the cooking is in progress or when cooking is com-
pleted, the timer signals will stop.
The use case diagram is illustrated for the common process of the cooker system as
shown in Fig. 2. This system consists of two main functionalities: Start Cooking and
End Cooking. The functional users of the input side of the cooker system are
DoorSensor, Start Button and the output sides of the functional users are Light and
Heater. Timer is on both sides of functional user.
Fig. 2. Use case diagram of Cooker System
3.2 COSMIC FSM in UML , SysML Sequence Diagram and Petri net
The measurement process of cooker system mainly comes from the basic functional
requirements. In this system, there are two functional processes of UML and SysML
sequence diagrams for cooker system as shown in Fig. 3 and 4. In Fig. 3, the cooker
checks that the door is open or closed. When the door is closed, it sends the signals to
heater, light and timer for one minute of cooking. The XML structure of UML and
SysML sequence diagram is extracted from Papyrus and several related tags to mes-
sage notation is identified and analyzed to calculate CFP. SysML borrowed from
UML are reused without modification. So, the concepts and features of SysML are the
16
T. Zaw, S. Z. Hlaing, M. M. Lwin, K. Ochimizu
same as UML. The tag is an important to calculate the data movements
based on COSMIC method. There are several types of message. In the XML struc-
ture, “Cooker Software” is defined as the boundary. If “messageSort” attribute in
tag is “asynchCall”, it will identify Entry and Exit data movements. If it is
synchronous call and move the data between the Cooker Software and other object, it
will identify Read and Write data movements. Entry data movement is defined by
extracting the value in “coveredBy” attribute in< lifeline> tag that is matched from
value in “receiveEvent” attribute in tag and value in “xmi:id” attribute in
fragment tag. Exit data movement is also defined by extracting the value in “cov-
eredBy” attribute in tag that is matched from value in “sendEvent” attribute
in tag and value in “xmi:id” attribute in tag.
Fig. 3. Start Cooking of UML and SysML Fig. 4. End Cooking of UML and SysML
The XML structure of Petri net is extracted from HiPS. Places are drawn as ellipses and held
multi-sets (bags) of tokens. A place models a local state given by its tokens. Transitions model
behaviors are drawn as boxes. A transition is connected to input places and output places by
arcs. Each data movement corresponds to the transition. The is a key to define the
data movements. The Petri net diagrams of cooker system are shown in Fig.5 and Fig.6.
Fig.5. Start Cooking in Petri net Fig. 6. End Cooking in Petri net
3.3 Mapping Phase
In this section, the key concepts of COSMIC are mapped to the different notations
such as: UML, SysML and Petri net. The mapping rules are described as follows:
Rule 1: The boundary represents in use case diagram. It shows the application border
that is established by identifying the external elements and application system.
Rule 2: The functional user is an active object that does not stop until either the com-
plete behaviour is executed or the object is terminated by some external object.
Rule 3: The functional process identifies use cases in the system.
Rule 4: The data groups identify the trigger event that carries data between objects.
Rule 5: The four data movements are identified as follows:
Rule 5.1: The Entry data movement identifies from Rule 2 to Rule 1.
17
IWSM/Mensura’18, September 18–20, 2018, Beijing, China
Rule 5.2: The Exit data movement identifies from Rule 1 to Rule 2.
Rule 5.3: The Read/Write data movement identify messages that send into or out of
the internal persistent storage.
Rule 6: Apply the COSMIC measurement function.
According to COSMIC measurement, each of the data movement in each functional
process is added to get the functional size of that process.
Rule 7: Aggregate the functional size measurements.
Aggregate all of the data movements of the functional processes of the whole system
into a single functional size value to obtain the functional size of the system.
3.4 Measurement Phase
After defining the mapping rules, the data movements of each functional process have
been identified as shown in Fig. 3 to 6. According to COSMIC standard, 1CFP is
defined as the size of one data movement. In the functional process of Cook End, it
has 1 Entry data movement that calculated the Stop attribute from timer. It also identi-
fied 2 Exit data movements which also counted the HeaterOff attribute to Heater, the
LightOff attribute to Light respectively. The subtotal of functional size for that func-
tional process is 3CFP. The total size from each function is 8CFP by adding all num-
ber of data movements. The data movement of each sequence diagram in this system
is as shown in Table 1.
Table 1. Measurement of data movements for cooker system
Process Message Functional User Data Movement CFP
Start PressSignal() Start button Entry 5 CFP
Cooking Getdoor-status(Close) Door Sensor Entry
HeatOn() Heater Exit
LightOn() Light Exit
StartCooking() Timer Exit
End Stop() Timer Entry 3
Cooking HeatOff() Heater Exit CFP
LightOff() Light Exit
Total Size 8 CFP
3.5 Automated Measurement for UML, SysML and Petri net
The prototype tool for the automatic measurement of the functions is developed as
an example in Java. The measurement result of the cooker system is shown in Fig7.
4 Evaluation
The automated function size measurement for cooker system is successfully im-
plemented in the three different notations. The manual measurement of FSM is time
consumption because different measurers need to measure the manual count of
18
T. Zaw, S. Z. Hlaing, M. M. Lwin, K. Ochimizu
Fig. 7. Automated Functional Size measurement results of Cooker System
software size correctly. But the measurers who are not experts use the automated
measurement of FSM to get more accurate size of software. The result of the manual
measurement of case study is the same result with the automated result.
5 Conclusion and Future Work
Software function size can be measured throughout the design phase in the software
development process. Like the other FSM methods, COSMIC FSM is the estimation
method for measuring the size of embedded software. In this paper, the proposed tool
applied the mapping rules between the XML structure from three different notations
and COSMIC FSM to estimate the functional size of software. The automated tool
contains the COSMIC rules for a simple case study of cooker system and helps to
estimate the effort in the early stages of development. It has been intended to propose
the large case studies by extending the COSMIC rules.
6 References
1. ISO 19761, Software Engineering – COSMIC : A Functional Size Measurement Method,
International Organization for Standardization, ISO, Geneva, 2011.
2. Symons, C.: Sizing and Estimating for Real-time Software – the COSMIC-FFP method.
In: DOD Software Tech News’, Editor: Data & Analysis Center for Software, USA DOD,
Rome NY, vol. 9(3), pp. 5–11 (2006).
3. Soubra, H., Abran, A. , Stern, S. , Ramdan-Cherif, A., “Design of a Functional Size Meas-
urement Procedure for Real-Time Embedded Software Requirements Expressed using the
Simulink Model”, IWSM-MENSURA, 2011.
4. Meiliana etal. ,“Automating Functional and Structural Software Size Measurement based
on XML Structure of UML Sequence Diagram ”, 2017 IEEE International Conference on
Cybernetics and Computational Intelligence 20-22 Nov. 2017.
5. Luigi Lavazza and Vieri Del Bianco, “A Case Study in COSMIC Functional Size Meas-
urement: the Rice Cooker Revisited”, IWSM/Mensura 2009.
6. A. Sellami. etal, “A measurement method for sizing the structure of UML sequence dia-
grams”, Information and Software Technology 59, 2015.
19