The Applicability of Present Estimation Models

to the Context of Mobile Applications

Laudson Silva de Souza and Gibeon Soares de Aquino Jr.

Department of Informatics and Applied Mathematics-DIMAp, Federal University of Rio Grande do Norte, Natal, Brazil

Keywords:

Software Engineering, Software Quality, Estimating Software, Systematic Review, Mobile Applications,

Mobile Computing.

Abstract:

The growing use of mobile technologies has shown different ways to access information and interact with

other computer systems. Thus, the traditional information systems are undergoing a process of adaptation

to this new computing environment. Thereafter, there is a need to reassess the current knowledge on the

planning and development of systems in this new environment. One area in particular that demand such

adaptation is the estimation software. The estimation processes, in general, are based on characteristics of the

systems, trying to quantify the complexity of implementing them. Hence, the main objective of this paper is

to present a proposal for an estimation model for mobile applications, and to debate about the applicability of

the traditional estimation models on this environment. Throughout the paper we analyze existing methods of

estimates, identify speciﬁc features of systems for mobile devices and ﬁnally an adaptation to be proposed for

this area of an existing estimation method.

1 INTRODUCTION

Computing is becoming increasingly present in peo-

ple’s lives and currently in a much more intense and

accelerated way due to the rise of the use of mobile

technologies in the world, such as mobile phones,

smartphones and tablets, all connected to mobile net-

works, which are increasingly more present in many

places and with better speeds. We are facing a new

technological scenario that is changing old habits and

creating new ways for the society to access informa-

tion and interact with computer systems (Naismith

et al., 2004), (Macario et al., 2009) and (Liu et al.,

2003).

The fact is that this new technological scenario

that is emerging with new requirements and restric-

tions requires a reevaluation of current knowledge

about the processes of planning and building software

systems. These new systems have different charac-

teristics and, therefore, an area in particular that de-

mands such adaptation is software estimation. The

estimation processes, in general, are based on char-

acteristics of the systems, trying to quantify the com-

plexity of implementing them. For this reason, it is

important to analyze the methods currently proposed

for software projects estimation and evaluate their ap-

plicability to this new context of mobile computing.

Hence, the main objective of this paper is to

present a proposal for an estimation model for mobile

applications, as well as discuss the applicability of tra-

ditional models used in estimation of information sys-

tems for the purpose of the development of systems

in the context of mobile computing. In this work, the

main estimation methods that exist now will be ana-

lyzed, the speciﬁc characteristics of mobile systems

will be identiﬁed and an adaptation of a estimation

method that exists in this context will be proposed.

2 MAIN ESTIMATION METHODS

In order to identify how the traditional estimation

methods could address the characteristics of the

systems, a literature review on the main estima-

tion methods was performed. The methods rec-

ognized by ISO identiﬁed in the survey can be

seen in Table 1. All methods with their descrip-

tions identiﬁed in the survey can be accessed at

http://www.laudson.com/methods.pdf.

Table 1 displays in chronological order the esti-

mation methods recognized by ISO, showing the year

of creation, the name of the method and the author

of it. At ﬁrst glance, one realizes that the main ex-

isting methods were not designed to consider the re-

249

Silva de Souza L. and Soares de Aquino Jr. G..

The Applicability of Present Estimation Models to the Context of Mobile Applications.

DOI: 10.5220/0004973002490254

In Proceedings of the 9th International Conference on Evaluation of Novel Approaches to Software Engineering (ENASE-2014), pages 249-254

ISBN: 978-989-758-030-7

 2014 SCITEPRESS (Science and Technology Publications, Lda.)

Table 1: Estimation Methods Recognized by ISO.

Year Method Author

1979 Function Point

Analysis (FPA)

Albrecht (Oligny

et al., 1999)

1988 Mark II FPA Charles Symons

(Symons, 2001)

1990 Netherlands

Software Metrics

Users Associa-

tion (NESMA)

FPA

The Netherlands

Software Metrics

Users Association

(Engelhart et al.,

2001)

1999 Common Soft-

ware Measure-

ment Interna-

tional Consor-

tium (COSMIC)

FFP

Common Soft-

ware Measurement

International Con-

sortium (COSMIC)

(Consortium et al.,

2007)

2004 Finnish Software

Metrics Associa-

tion FSM

The Finnish Soft-

ware Metrics As-

sociation (FiSMA)

(Forselius, 2004)

quirements of mobile applications. Indeed, the very

creation of most of them precedes the emergence of

mobile devices as we know today. This suggests that

the use of these methods to estimate the effort of the

development of projects involving systems or applica-

tions for mobile devices would cause a possible fail-

ure to quantify the complexity of some features and,

therefore, would not produce adequate estimates.

3 CHARACTERISTICS OF

MOBILE APPLICATIONS

In order to identify characteristics that are inherent to

systems and mobile applications, a surveying of the

characteristics of these types of software was accom-

plished through a systematic review. Conducting a

systematic review is relevant because most searches

begin with some kind of review of the literature, and

a systematic review summarizes the existing work

fairly, without inclinations. So the surveys were con-

ducted according to a predeﬁned search strategy, in

which the search strategy should allow the integrity

of the research to be evaluated. The planning and ac-

complishment of the methodology discussed were di-

rected by Procedures for Performing Systematic Re-

views (Kitchenham, 2004).

3.1 Systematic Review

In the context of research questions, the following re-

search question was formulated: “What are the char-

acteristics of Mobile Applications?”, based on the is-

sue about the proposed study.

Procedures for The Evaluation of the Articles: the

articles will be analyzed considering its relation with

the issues addressed in the research questions, inclu-

sion criteria and exclusion criteria, and their respec-

tive situation will be assigned with either ”Accepted”

or ”Rejected”. The evaluation will follow the follow-

ing procedure: read the title and abstract and, should

it be related with the research question, also read the

whole article. The implementation of the systematic

review was performed almost in line with its plan-

ning, except for the need to adjust the syntax of the

proposed search string due to the particularities of

the research bases. 234 articles were analyzed, of

which 40 were selected and considered “Accepted”

according to the inclusion criteria; 194 were consid-

ered “Rejected” according to the exclusion criteria.

The list with all the articles can be accessed at the

following address: http://http://www.laudson.com/sr-

articles.pdf.

Given the results extracted from the systematic re-

view, it’s is possible to identify 29 kinds of character-

istics in 100% of the articles evaluated and consid-

ered accepted in accordance with the inclusion crite-

ria. However some of these are a mixture of character-

istics of mobile devices and characteristics of mobile

applications, such as the characteristic called “Lim-

ited Energy”, which is a characteristic of the device

and not the application, however the articles that men-

tion this type of characteristic emphasize that in the

development of a mobile application, this “limitation”

must be taken into account since all the mobile de-

vices are powered by batteries, which have a limited

life, depending completely on what the user operates

daily. Applications requiring more hardware or soft-

ware resources will consume more energy. The 23

types of characteristics mentioned the most in the se-

lected articles can be observed following. There is a

description of each characteristic identiﬁed in the re-

view:

◦ Limited energy (Sohn et al., 2005); Small screen

(Sohn et al., 2005); Limited performance (Mukhtar

et al., 2008); Bandwidth (Mukhtar et al., 2008);

Change of context (Mukhtar et al., 2008); Reduced

memory (Sohn et al., 2005); Connectivity (Feng,

2009); Interactivity (Mukhtar et al., 2008); Storage

(Mukhtar et al., 2008); Software portability (Mukhtar

et al., 2008); Hardware portability (Mukhtar et al.,

2008); Usability (Feng, 2009); 24/7 availability

ENASE2014-9thInternationalConferenceonEvaluationofNovelSoftwareApproachestoSoftwareEngineering

250

(Feng, 2009); Security (Feng, 2009); Reliability (Ku-

mar Maji et al., 2010); Efﬁciency (Al-Jaroodi et al.,

2009); Native vs. Web Mobile (Feng, 2009); In-

teroperability (Mukhtar et al., 2008); Response time

(Hameed et al., 2010); Privacy (Feng, 2009); Short

term activities (Alekhya Mandadi, 2009); Data in-

tegrity (Hameed et al., 2010); Key characteristics

(Alekhya Mandadi, 2009); Complex integration of

real-time tasks (Hayenga et al., 2008); Constant inter-

ruption of activities (Alekhya Mandadi, 2009); Func-

tional area (Giessmann et al., 2012); Price (Giess-

mann et al., 2012); Target audience (Giessmann et al.,

2012); Provider type (Giessmann et al., 2012).

4 CHARACTERISTICS OF

MOBILE APPLICATIONS

SURVEY

With the conclusion of the systematic review, a sur-

vey was carried out among experts in mobile devel-

opment with the purpose of ratifying the character-

istics previously raised and to prove their respective

inﬂuence on mobile development. The disclosure of

the survey was conducted in more than 70 locations,

among them universities and businesses, through e-

mails, study groups and social groups. In general, of

all 117 feedbacks received through the survey, 100%

of the experts conﬁrmed the characteristics; among

them, an average of 72% indicated a greater effort and

complexity regarding the characteristics during devel-

opment, an average of 12% indicated less effort and

complexity and, ﬁnally, an average of 16% indicated

they did not perceive any difference in mobile devel-

opment, even though they conﬁrmed the presence of

the characteristics.

5 PROBLEM ADDRESSED

As noted in Section II, there is no estimation method

developed for mobile applications projects. More-

over, some of the characteristics elicited in Section

III aggravate the complexity and, thereafter, the effort

in the development of mobile applications. From the

analysis that follows, with the characteristics of ap-

plications on mobile devices elicited in Section III,

it is clear that they are different from the charac-

teristics of traditional systems and directly inﬂuence

its development. A clear example, which is differ-

ent from the information or desktop systems, is the

characteristic that the mobile devices have “Limited

Energy”. As mobile devices are powered by battery,

which have a limited lifetime period, the applications

must be programmed to require the minimal amount

of hardware resources possible, since the more re-

sources consumed, the greater amount of energy ex-

pended. This characteristic makes it necessary for the

solution project to address this concern, generating a

higher complexity of development and, thereafter, a

greater effort and cost.

Another speciﬁc characteristic of this context is

the “Graphical Interface”. Due to the reduced screen

size, the interface design is limited. Therefore, a

greater complexity and, thereafter, a larger effort is re-

quired in the development of the graphical interface.

Another characteristic related to the screen is the “In-

put Interface”, which deﬁnes how the user will inter-

act with the application, in other words, if the user

will interact via keypad, stylus, touch screen or voice

and image recognition. The latter makes the task of

developing applications that offers all these interac-

tion options more complex, thus requiring a bigger

effort.

Regarding connectivity, the characteristic “Band-

width” was identiﬁed, wherein a mobile application

might have the maximum band at times and the min-

imum in other moments. Some types of applications

need to realize this and act differently in each sit-

uation. Another related feature is the “Connectiv-

ity Type”. Mobile applications can be developed to

support different types of connectivity such as 3G,

bluetooth, infrared, Wi-Fi, Wireless, NFC and others.

In addition, a single application can support multiple

types of connectivity simultaneously. These behav-

iors directly affect the complexity of the software and

therefore require a larger development effort.

The “Change in Context” is also another charac-

teristic inherent in mobile applications, which should

take into account not only the data entries explicitly

provided by users, but also the implicit entries con-

cerning the physical and computational context of the

users and the environments that surround them. In

addition, the “Constant Interruption of Activities” is

a much more common characteristic in this context,

as well as the need for some applications to be devel-

oped to work ofﬂine and therefore be able to synchro-

nize. Mobile applications should be prepared for dif-

ferent scenarios because the activities are interrupted

constantly. Receiving a call, lack of connection and

low battery are examples of such interruptions, which

makes the applications become much more complex.

Despite the advances related to the computational

ability of these devices, their hardware must still be

considered as limited, especially when compared to

desktops and servers. Two characteristics related to

this issue are “Limited Performance” and “Reduced

TheApplicabilityofPresentEstimationModelstotheContextofMobileApplications

251

Memory”. Besides these, a characteristic inherent to

the use of mobile devices is the “Response Time”,

that is directly related to the power of “Processing”.

Mobile applications must be initialized and ﬁnalized

immediately, in other words, any development should

be focused in the time variable. These characteristics

require the applications to be developed with a pos-

sible resource optimization for a better efﬁciency and

response time, requiring more effort.

The “Portability” is also a required characteristic

of these applications. It can be divided into two char-

acteristics: the “Hardware Portability” and the “Soft-

ware Portability”. Regarding the ﬁrst one, nowadays

there is a large number of different devices with dif-

ferent capabilities and resources. A mobile applica-

tion should be able to run on the largest number of

devices possible. This requires an increased effort in

the development. Moreover, a greater effort in testing

this kind of portability is required. Regarding “Soft-

ware Portability”, it is necessary to develop speciﬁc

applications for each existing platform should the ap-

plication be native. With this, more effort is required

for replications of the same software product, includ-

ing the tests.

Finally, mobile applications can be separated into

two types: Native or Web Mobile. The ﬁrst one

has higher performance and easiness in accessing the

hardware, while the second has lower performance

since it is web, but it is easier to achieve portability. In

addition, there are some applications that are consid-

ered hybrids. Depending on the type of application,

the issues that must be considered and the complexity

can be different, requiring different development ef-

forts.

From the survey of the most popular estimation

methods cited in Section III, it was found that these

characteristics are not covered by the current estima-

tion methods for two explicit reasons: ﬁrst, none of

the existing methods was designed to perform project

estimation in mobile applications development; and

second, all the characteristics discussed in this sec-

tion are exclusive to mobile applications, with direct

interference in their development, thereby generating

a greater complexity and, thereafter, a greater effort.

However, to consider any of the existing estimation

methods to apply to the process of development of

mobile applications is to assume that this kind of de-

velopment is no different than the project of develop-

ing desktop applications, in other words, an eminent

risk is assumed.

6 PROPOSAL: ESTIMATION IN

MOBILE APPLICATION

DEVELOPMENT PROJECT

The approached proposed is an adaptation of an ex-

isting method, based exclusively on methods recog-

nized as international standards by ISO. Among the

most popular estimation methods mentioned in Sec-

tion III, the method used to base the proposal below

on is known as “Finnish Software Metrics Associa-

tion (FISMA)”. The model is one of the ﬁve meth-

ods for measuring software that complies with the

ISO/IEC 14143-1 standard, is accepted as an inter-

national standard for software measuring (Forselius,

2004) and nowadays over 750 software projects are

completed being estimated by FISMA. However, the

difference between this and other methods that are in

accordance with the above standard, which are the

Common Software Measurement International Con-

sortium Function Points (COSMIC FP) (Consortium

et al., 2007), the International Function Point Users

Group (IFPUG) FPA (Oligny et al., 1999), MarkII

FPA (Symons, 2001) and the Netherlands Software

Metrics Association (NESMA) WSF (Gencel et al., ),

is that the method used is based in functionality but

is service-oriented. It also proposes in its deﬁnition

that it can be applied to all types of software, but this

statement is lightly wrong since in its application, the

method does not take into account the characteristics

elicited in Section IV.

Overall, the FISMA method proposes that all

services provided by the application are identiﬁed.

It previously deﬁnes some services, among which

stands out the user’s interactive navigation, consulting

services, user input interactive services, interface ser-

vices for other applications, data storage services, al-

gorithmic services and handling services. Finally, af-

ter identifying all the services, the size of each service

is calculated using the same method and thus obtain-

ing a total functional size of the application by adding

the size of each service found (FiSMA, 2004).

6.1 Applying the Chosen Model

The FiSMA method can be applied manually or with

the aid of the Experience Service

tool, which was the

case, provided by FiSMA itself through contact made

with senior consultant Pekka Forselius and with the

chairman of the board Hannu Lappalainen.

When using the tool, it is necessary to perform all

the steps of the previous subsection to obtain the func-

tional size. Figure 1 shows the ﬁnal report after the

http://www.experiencesaas.com/

ENASE2014-9thInternationalConferenceonEvaluationofNovelSoftwareApproachestoSoftwareEngineering

252

implementation of the FiSMA on a real system, the

Management of Academic Activities Integrated Sys-

tem (Sigaa) in its Mobile version, developed by the

Superintendence of Computing (SINFO) of the Fed-

eral University of Rio Grande do Norte (UFRN).

Figure 1: Final Report of FiSMA applied to Sigaa Mobile.

After the application of FiSMA, the functional

size of the software is obtained and from this it is pos-

sible to ﬁnd the effort using the formula: Estimated

effort (h) = size (fp) x reuse x rate of delivery (h/fp) x

project status; the latter is related to productivity fac-

tors that are taken into account for the calculation of

the effort. However, of the factors predeﬁned by the

FiSMA regarding the product, only 6 (six) are pro-

posed, in which the basic idea of the evaluation is that

“the better the circumstances of the project, the more

positive the assessment”. The weighting goes from -

- to + +, as follows: Caption: ◦ (+ +) = [1.10] Excel-

lent situation, much better circumstances than in the

average case; ◦ (+) = [1.05] Good situation, better cir-

cumstances than in the average case; ◦ (+ / -) = [1.0]

Normal situation; ◦ (-) = [0.95] Bad situation, worse

circumstances than in the average case; ◦ (- -) = [0.90]

Very bad situation, much worse circumstances than in

the average case.

Among the productivity factors mentioned above,

only the “Portability Requirement” factor ﬁts in har-

mony with the “Portability” characteristic regarding

both hardware and software. However, none of the

other factors discusses the characteristics of mobile

application, in other words, after obtaining the func-

tional size of the software and applying the productiv-

ity factors related to the product to estimate the effort,

this estimate ignores all of the characteristics of mo-

bile applications, judging that the estimate of tradi-

tional information systems is equal to the mobile ap-

plication. However, with the proposal of the creation

of new productivity factors, which would be the spe-

ciﬁc characteristics of mobile applications, this prob-

lem will be solved, as presented below.

Performance Factor: ◦ (-) The application should

be concerned with the optimization of resources for a

better efﬁciency and response time. ◦ (+/-) Resource

optimization for better efﬁciency and response time

may or may not exist. ◦ (+) Resource optimization

for better efﬁciency and response time should not be

taken into consideration.

Power Factor: ◦ (-) The application should be con-

cerned with the optimization of resources for a lower

battery consumption. ◦ (+/-) Resource optimization

for lower battery consumption may or may not exist.

◦ (+) Resource optimization for a lower battery con-

sumption should not be taken into consideration.

Band Factor: ◦ (-) The application shall require

the maximum bandwidth. ◦ (+/-) The application

shall require reasonable bandwidth. ◦ (+) The appli-

cation shall require a minimum bandwidth.

Connectivity Factor: ◦ (-) The application must

have the maximum willingness to use connections

such as 3G, Wi-ﬁ, Wireless, Bluetooth, Infrared and

others. ◦ (+/-) The application must have reasonable

predisposition to use connections such as 3G, Wi-Fi

and Wireless. ◦ (+) The application must have only a

predisposition to use connections, which can be: 3G,

Wi-ﬁ, Wireless, Bluetooth, Infrared or others.

Context Factor: ◦ (-) The application should work

ofﬂine and synchronize. ◦ (+/-) The application

should work ofﬂine and it is not necessary to synchro-

nize. ◦ (+) The application should not work ofﬂine.

Graphic Interface Factor: ◦ (-) The application has

limitations due to the screen size because it will be

mainly used by cell phone users. ◦ (+/-) The applica-

tion has reasonable limitation due to the screen size

because it will be used both by cell phone and tablet

users. ◦ (+) The application has little limitation due

to the screen size because it will be mainly used by

tablet users.

Input Interface Factor: ◦ (-) The application must

have input interfaces for touch screen, voice, video,

keyboard and others. ◦ (+/-) The application must

have standard input interfaces for keyboard. ◦ (+) The

application must have any one of the types of inter-

faces, such as: touch screen, voice, video, keyboard

or others.

The proposed factors take into account the same

weighting proposed by FiSMA, but only ranging from

- to +, in other words: ◦ (+) = [1.05] Good situation,

better circumstances than in the average case; ◦ (+ /

-) = [1.0] Normal Situation; ◦ (-) = [0.95] Bad situ-

ation, worse circumstances than in the average case.

The functional size remains the same, thus affecting

only the formula used to obtain the effort, which will

now consider in its “project situation” variable the

new productivity factors speciﬁc for mobile applica-

tions.

TheApplicabilityofPresentEstimationModelstotheContextofMobileApplications

253

7 CONCLUSIONS

Given the results presented, based on the literature re-

view of estimation methods and on the systematic re-

view of the characteristics of mobile applications, it

was observed that this sub-area of software engineer-

ing still falls short. Basically, it’s risky to use any ex-

isting estimation method in development projects for

mobile applications, as much as there are some mod-

els already widespread in industry, such as the Func-

tion Point Analysis, the Mark II and the COSMIC-

FFP, which are even approved by ISO as international

standards. They all fall short by not taking into ac-

count the particularities of mobile applications, which

makes the method partially ineffective in this situa-

tion.

The validation shall be as follows, will be raised

the total effort expended in developing the Sigaa Mo-

bile project. After the method is applied to estimate

FISMA, in his original proposal thus obtaining an es-

timate of effort. Then the proposed adjustment will

be applied also generating an effort estimate ﬁnally

the comparative analysis between the three estimates

generated will be performed to determine which pro-

posal is closer to the actual effort spent.

Based on this study, it is concluded that the pro-

posal presented in this work is entirely appropriate

and viable and that this proposal should take into ac-

count all the peculiarities of such applications, ﬁnally

creating a belief that there actually are considerable

differences in the development project for mobile ap-

plications.

REFERENCES

Al-Jaroodi, J., Al-Dhaheri, A., Al-Abdouli, F., and Mo-

hamed, N. (2009). A survey of security middleware

for pervasive and ubiquitous systems. In Network-

Based Information Systems, 2009. NBIS ’09. Interna-

tional Conference on, pages 188–193. IEEE.

Alekhya Mandadi, Deepti Mudegowder, I. R. D. E. A. G.

(2009). Mobile applications: Characteristics & group

project summary. Mobile Application Development.

Google.

Consortium, C.-C. S. M. I. et al. (2007). The cosmic func-

tional size measurement method-version 3.0 measure-

ment manual (the cosmic implementation guide for

iso/iec 19761: 2003).

Engelhart, J., Langbroek, P., et al. (2001). Function Point

Analysis (FPA) for Software Enhancement. NESMA.

Feng, H. (2009). A literature analysis on the adoption of

mobile commerce. In Grey Systems and Intelligent

Services, 2009. GSIS 2009. IEEE International Con-

ference on, pages 1353–1358. IEEE.

FiSMA, F. S. M. A. (2004). Fisma functional size measure-

ment method version 1-1.

Forselius, P. (2004). Finnish software measurement asso-

ciation (ﬁsma), fsm working group: Fisma functional

size measurement method v. 1.1.

Gencel, C., Heldal, R., and Lind, K. On the conversion be-

tween the sizes of software products in the life cycle.

Giessmann, A., Stanoevska-Slabeva, K., and de Visser, B.

(2012). Mobile enterprise applications–current state

and future directions. In System Science (HICSS),

2012 45th Hawaii International Conference on, pages

1363–1372. Google.

Hameed, K. et al. (2010). Mobile applications and systems.

Google.

Hayenga, M., Sudanthi, C., Ghosh, M., Ramrakhyani, P.,

and Paver, N. (2008). Accurate system-level perfor-

mance modeling and workload characterization for

mobile internet devices. In Proceedings of the 9th

workshop on MEmory performance: DEaling with

Applications, systems and architecture, MEDEA ’08,

pages 54–60, New York, NY, USA. ACM. ACM.

Kitchenham, B. (2004). Procedures for performing system-

atic reviews. Keele, UK, Keele University, 33:2004.

Kumar Maji, A., Hao, K., Sultana, S., and Bagchi, S.

(2010). Characterizing failures in mobile oses: A case

study with android and symbian. In Software Relia-

bility Engineering (ISSRE), 2010 IEEE 21st Interna-

tional Symposium on, pages 249–258. IEEE.

Liu, T., Wang, H., Liang, J., Chan, T.-W., Ko, H., and Yang,

J. (2003). Wireless and mobile technologies to en-

hance teaching and learning. Journal of Computer As-

sisted Learning, 19(3):371–382.

Macario, G., Torchiano, M., and Violante, M. (2009). An

in-vehicle infotainment software architecture based on

google android. In Industrial Embedded Systems,

2009. SIES ’09. IEEE International Symposium on,

pages 257–260.

Mukhtar, H., Bela

ıd, D., and Bernard, G. (2008). A model

for resource speciﬁcation in mobile services. In Pro-

ceedings of the 3rd international workshop on Ser-

vices integration in pervasive environments, SIPE ’08,

pages 37–42, New York, NY, USA. ACM. ACM.

Naismith, L., Sharples, M., Vavoula, G., Lonsdale, P., et al.

(2004). Literature review in mobile technologies and

learning.

Oligny, S., Desharnais, J.-M., and Abran, A. (1999). A

method for measuring the functional size of embed-

ded software. In 3rd International Conference on In-

dustrial Automation, pages 7–9.

Sohn, J.-H., Woo, J.-H., Lee, M.-W., Kim, H.-J., Woo,

R., and Yoo, H.-J. (2005). A 50 mvertices/s graph-

ics processor with ﬁxed-point programmable vertex

shader for mobile applications. In Solid-State Cir-

cuits Conference, 2005. Digest of Technical Papers.

ISSCC. 2005 IEEE International, pages 192–592 Vol.

1. Google.

Symons, C. (2001). Come back function point analysis

(modernized)–all is forgiven!). In Proc. of the 4th Eu-

ropean Conference on Software Measurement and ICT

Control, FESMA-DASMA, pages 413–426.

ENASE2014-9thInternationalConferenceonEvaluationofNovelSoftwareApproachestoSoftwareEngineering

254