SSN2225-
658X(
ONLINE)
I
SSN2412-
6551(
PRI
NT)
Vol
ume7,I
ssue1
2017
International Journal of
DIGITAL INFORMATION AND WIRELESS COMMUNICATIONS
Editor-in-Chief Overview
The SDIWC International Journal of Digital Information and Wireless
Communications is a refereed online journal designed to address the
Dr. Rohaya Latip, University Putra Malaysia, Malaysia networking community from both academia and industry, to discuss
recent advances in the broad and quickly-evolving fields of computer and
Editorial Board communication networks, technology futures, national policies and
standards and to highlight key issues, identify trends, and develop visions
Abbas M. Al-Ghaili, Universiti Putra Malaysia, Malaysia
for the digital information domain.
Ali Sher, American University of Ras Al Khaimah, UAE
Altaf Mukati, Bahria University, Pakistan In the field of Wireless communications; the topics include: Antenna
Andre Leon S. Gradvohl, State University of Campinas, Brazil systems and design, Channel Modeling and Propagation, Coding for
Azizah Abd Manaf, Universiti Teknologi Malaysia, Malaysia Wireless Systems, Multiuser and Multiple Access Schemes, Optical
Bestoun Ahmed, University Sains Malaysia, Malaysia Wireless Communications, Resource Allocation over Wireless Networks,
Carl Latino, Oklahoma State University, USA Security, Authentication and Cryptography for Wireless Networks, Signal
Dariusz Jacek Jakbczak, Technical University of Koszalin, Poland
Processing Techniques and Tools, Software and Cognitive Radio, Wireless
Duc T. Pham, University of Bermingham, UK
E.George Dharma Prakash Raj, Bharathidasan University, India
Traffic and Routing Ad-hoc networks, and Wireless system architectures
Elboukhari Mohamed, University Mohamed First, Morocco and applications. As one of the most important aims of this journal is to
Eric Atwell, University of Leeds, United Kingdom increase the usage and impact of knowledge as well as increasing the
Eyas El-Qawasmeh, King Saud University, Saudi Arabia visibility and ease of use of scientific materials, IJDIWC does NOT CHARGE
Ezendu Ariwa, London Metropolitan University, United Kingdom authors for any publication fee for online publishing of their materials in
Fouzi Harrag, UFAS University, Algeria the journal and does NOT CHARGE readers or their institutions for
Genge Bela, University of Targu Mures, Romania accessing to the published materials.
Guo Bin, Institute Telecom & Management SudParis, France
Hadj Hamma Tadjine, Technical university of Clausthal, Germany Publisher
Hassan Moradi, Qualcomm Inc., USA The Society of Digital Information and Wireless Communications
Isamu Shioya, Hosei University, Japan 20/F, Tower 5, China Hong Kong City, 33 Canton Road, Tsim Sha Tsui,
Jacek Stando, Lodz University of Technology, Poland
Kowloon, Hong Kong
Jan Platos, VSB-Technical University of Ostrava, Czech Republic
Jose Filho, University of Grenoble, France Further Information
Juan Martinez, Gran Mariscal de Ayacucho University, Venezuela
Website: http://sdiwc.net/ijdiwc, Email: ijdiwc@sdiwc.net,
Kaikai Xu, University of Electronic Science and Technology of China, China
Khaled A. Mahdi, Kuwait University, Kuwait Tel.: (202)-657-4603 - Inside USA; 001(202)-657-4603 - Outside USA.
Ladislav Burita, University of Defence, Czech Republic Permissions
Maitham Safar, Kuwait University, Kuwait
International Journal of Digital Information and Wireless Communications
Majid Haghparast, Islamic Azad University, Shahre-Rey Branch, Iran
Martin J. Dudziak, Stratford University, USA
(IJDIWC) is an open access journal which means that all content is freely
Mirel Cosulschi, University of Craiova, Romania available without charge to the user or his/her institution. Users are
Mohamed Amine Ferrag, Guelma University, Algeria allowed to read, download, copy, distribute, print, search, or link to the
Monica Vladoiu, PG University of Ploiesti, Romania full texts of the articles in this journal without asking prior permission
Nan Zhang, George Washington University, USA from the publisher or the author. This is in accordance with the BOAI
Noraziah Ahmad, Universiti Malaysia Pahang, Malaysia definition of open access.
Pasquale De Meo, University of Applied Sciences of Porto, Italy
Paulino Leite da Silva, ISCAP-IPP University, Portugal
Disclaimer
Piet Kommers, University of Twente, The Netherlands
Radhamani Govindaraju, Damodaran College of Science, India
Statements of fact and opinion in the articles in the International Journal
Ramadan Elaiess, University of Benghazi, Libya of Digital Information and Wireless Communications (IJDIWC) are those
Rasheed Al-Zharni, King Saud University, Saudi Arabia of the respective authors and contributors and not of the International
Su Wu-Chen, Kaohsiung Chang Gung Memorial Hospital, Taiwan Journal of Digital Information and Wireless Communications (IJDIWC) or
Talib Mohammad, University of Botswana, Botswana The Society of Digital Information and Wireless Communications (SDIWC).
Tutut Herawan, University Malaysia Pahang, Malaysia Neither The Society of Digital Information and Wireless Communications
Velayutham Pavanasam, Adhiparasakthi Engineering College, India nor International Journal of Digital Information and Wireless
Viacheslav Wolfengagen, JurInfoR-MSU Institute, Russia
Communications (IJDIWC) make any representation, express or implied,
Wen-Tsai Sung, National Chin-Yi University of Technology, Taiwan
Wojciech Zabierowski, Technical University of Lodz, Poland
in respect of the accuracy of the material in this journal and cannot accept
Yasin Kabalci, Nigde University, Turkey any legal responsibility or liability as to the errors or omissions that may
Yoshiro Imai, Kagawa University, Japan be made. The reader should make his/her own evaluation as to the
Zanifa Omary, Dublin Institute of Technology, Ireland appropriateness or otherwise of any experimental technique described.
Zuqing Zhu, University of Science and Technology of China, China
Copyright 2017 sdiwc.net, All Rights Reserved
TABLE OF CONTENTS
Original Articles
COMPARATIVE ANALYSES ON LOGO IMAGE DESIGNS BETWEEN ARAB Ahmad Eibo, Toshiyuki
1
AND JAPAN Yamashita, Keiko Kasamatsu
ENERGY EXPENDITURE PER BIT MINIMIZATION INTO MB-OFDM UWB Houda Chihi, Reza Zaeem, Ridha
42
SYSTEMS Bouallegue
A SURVEY OF FOUR ROUTING PROTOCOLS FOR WIRELESS SENSOR Raid Zaghal, Saeed Salah, Fadi
49
NETWORKS Alyounis
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 1-9
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
1
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 1-9
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
since Japan has adopted English in compulsory local Japan and Arab market as shown in Figure 1.
education even daily life Japanese people have
become much more knowledgeable of English
language alphabets than other foreign languages
[7]. The need for appropriate image of Arabic
alphabetic logotype design to influence Japanese
people positively is compelling.
3 PURPOSE
2
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 1-9
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
revealed three groups of substantially correlated pair was considered a main factor and respectively
variables representing three main factors with represented the level of the following images: 1)
given values equals or more than 1 ( 1) and "Familiar- Favorable", 2) "Creative-Innovative" ,
cumulative contribution rate of 62.514%. The and 3) "Traditional-Consistent" used in logos
following three variables combinations "3. design. For the sake of simplicity, we employed
Familiar" and "6. Favorable", "10. Creative" and "2. Varimax rotation technique to determine the
Innovative ", and "7. Traditional" and "4. dimensionality of extracted factors and their
Consistent" have high factor loadings scores, each relationships as in Table 1 below.
3
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 1-9
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
Table 2. Cluster
The examination number of cases in the final reveals that in spite of that Familiar-Favorable
cluster solution in Table 3 shows that average image level for Arabic logotypes is relatively lower
scores of main factor "Favorable-Familiar" and than is for Japanese logotypes (-0.698*<0.499*),
Creative-Innovative is significant for Japan Creative-Innovative image level is, on the
group (J) and Arab group (A). Analytic comparison contrary, greatly higher for Arabic logotypes than
between average scores of two groups (J) and (A) in Japanese logotypes (0.251**>-0.179**).
% of % of % of % of % of
Case N
Combined cases Total cases F1 F2 F3
Total 24 100.0%
4
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 1-9
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
The results of clusters analyses were consistent Because we found that Arabic logotypes are
with prior works [12] this is especially true for unfamiliar by Japanese people, we can now test
"Familiar-Favorable", "Creative-Innovative and whether or not this result is the same for English
"Traditional-Consistent" main images, where the logotypes. Since the first survey examined Arabic
patterns of the path estimates are consistent across and Japanese logotypes, our interest in the second
response variables and obtained clusters. But the one was focused on Arabic and English logotypes.
effect of unfamiliarity & creativity on participants 50 undergraduate students (35 males and 15
was higher for Arabic logotypes than for Japanese females) participated in this survey. We applied the
logotypes. Our findings reveal that the perception same questionnaire structure used in survey I for 20
of Arabic and Japanese logotypes by Japanese well-known logos of 10 global corporate in Arab
people towards the same kind of product differs and Japan with Arabic and English alphabetic
because of the unfamiliarity with Arabic language logotypes respectively: FedEx, Burger King, CNN,
alphabets and familiarity with the Japanese Subway, Tide, Baskin Robbin, Vodafone,
characters, however, a strong image of creativity Starbucks, Coca Cola, and Adidas as is shown in
was found in Arabic logo design attributes Figure 4.
although Arabic alphabets are not well-known in
Japan. Would our finding be the same for English
logos design? In other words, the question of
whether or not English alphabetic logotypes would
have about the same effect of Arab logos design on
Japanese people is worth asking. A study done by
Takahashi [13] showed that although Japanese
graduate students do not often understand English,
they do not have negative image of it. The
sociocultural situation in Japanese is truly unique
due to American culture influence. The possible
reasons for the extensive filtering of American
culture into the Japan are manifold. In Japan,
generally Western languages were seen as
symbolic of progress and modernization. After
World War II Japanese aspired to creating a
country similar in economic and technological
progression to the United States of America.
Through desire to emulate the American way of
life, the Japanese people began, willingly, to adopt
much more American culture [14] In recent
decades, even though Japan has certainly reached Figure 2. Examples of corporate logos
its goal of becoming a modern, economically
powerful society, America still maintain its allure 5. 1 Factor Analyses
to the Japanese as a culture which is fashionable,
and generally appealing. This is not simply a matter In survey II, factor analyses also prioritize three
of cultural contact, but part of a complex process of significant factors resembling main factors in
identity reformation mediated by a sense of survey I 3. Familiar and 6. Favorable, 10.
tendency related to the representation and Creative and 2. Innovative, and 7. Traditional
appropriation of the other. Despite the evident and 4. Consistent as is shown in Table 4.
importance of research on such a wide-ranging and Astonishingly, the three influential images of
complex phenomenon of cultural and differences, corporate and products logotypes are the same in
hardly any attempt has been made until recently, both surveys as it has been found in relevant
hence in this paper we wanted to include studying research conducted previously [11, 16] In Table 4.
the significance of English logotype influence [15] Since the result is relatively consistent with what
5
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 1-9
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
survey I concluded, it fairly suggest that the designed based on specific set between the three
association made between three logo images of seems to influence Japanese people perception, and
Familiar-Favorable, Creative-Innovative, and therefore it is recommended to be considered as
Traditional-Consistent is highly significant. main images in logos design for products and
Thus, the combination of logo structures are being corporate.
the average Scores of factor 1 (F1) and factor 2 In the contrast to what we found in Figure 3, the
(F2) in Figure 3 shows that the image of average Scores of factor 2 (F2) and factor 4 (F3) in
Familiar-Favorable scored low for Arabic Figure 4 reveals that the image of
logotypes and high for English logotypes although Innovative-Creative scored high in Arabic
logo designed marks are similar in shapes and logotypes than in English logotypes. It clarifies
colors relatively, indicating presence of unfamiliar high level of creative attribute towards Arabic
attribute towards Arabic logotypes. logotypes.
Figure 3. Factor 2 and Factor 3Average scores Figure 4. Factor 2 and Factor 3Average scores
6
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 1-9
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
5.2 Rough Sets Analyses decision table below between logo design structure
(logo mark, logo entirety, and logotype) and main
Although Factor Analyses extracted three logo three images (Traditional-Consistent,
dimensions of the main corporate logo images Familiar-Favorable, and Innovative-Creative) is
"Tradition-Consistent", "Favorable-Familiar", and shown in Table 5: 1) Target set U, 2) Attribute Set
Creative-Innovative, how logo attributes Condition C, and 3) Attribute Set Decision D.
influence consumer perception was still unclear. In Target set U is made of corporate logo, Attribute
order to obtain adequate contraction of minimum Set Condition C consists of the following
attributes combination as well as identifying the equations: C = {Logo Mark (1 for Present, 0 for
psychological relationship between logo samples Absent), Active-Inactive (1 for Active, 0 for
and consumer perception, an additional inactive), Solid sense (1 for Present, 0 for Absent),
mathematical analyses approach known as Rough Logotype thickness (1 for thin, 0 for thick),
Sets proposed by Zdzisaw Pawlak [17] would be Logotype Italic sense (1 for Present, 0 for Absent),
imperative. In psychology, it is been used widely Logotype variation of base line (1 for Present, 0 for
for data explicit interpretation and accurate Absent), Logotype white color processing of (1 for
minimal sets by revealing the causal relationship Present, 0 for Absent), Logotype language type (1
between If and Then rule decisions [16, 18, 19, for Known, 0 for Unknown}, and finally Attribute
20] Rough Set analyses require binary variables Set Decision D of three main images of corporate
for a case or an event. Each variable takes the logo D = {Traditional-Consistent and
values 0 or 1, that is, each case or event is Familiar-Favorable, and Creative-Innovative} with
conceived as a configuration of conditions. Data 1 given value when image score is greater than zero
have the form of a decision table in which the and 2 value when image score is lower than zero.
columns represent causal variables (logical At this stage of data preparations, the number of
variables) may take the value 0 or 1and the rows attribute decision sets can be minimized by logical
represent cases [21] In Table 5 the composition contraction.
Table 5. Decision Table
J1 0 0 1 0 0 0 0 1 1 1 2
A1 0 0 0 0 0 0 0 0 2 2 1
J2 1 1 0 0 1 0 0 1 2 1 2
A2 1 1 0 0 1 0 0 0 2 2 1
J3 0 0 0 0 0 0 1 1 2 1 2
A3 0 0 1 0 0 0 1 0 2 2 2
J4 0 1 0 0 1 0 1 1 2 1 1
A4 0 1 1 0 0 0 1 0 2 2 2
J5 1 0 0 0 1 0 0 1 2 1 1
A5 1 0 1 0 1 0 1 0 2 1 1
J6 0 0 0 0 0 0 0 1 2 1 2
A6 0 0 1 0 1 0 0 0 2 2 2
J7 1 0 0 0 0 0 0 1 1 1 1
A7 1 1 1 0 0 0 1 0 2 2 2
J8 1 0 1 1 2 1
A8 1 0 1 1 2 1
J9 1 1 1 1 1 0 1 1 1 1 1
A9 0 1 1 1 1 0 1 0 2 2 1
J10 1 0 1 0 0 0 0 1 1 1 1
A10 1 0 1 1 0 0 1 0 1 2 2
7
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 1-9
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
The goal of rough ret analyses is to specify the And [Italic Sense Present] And [White Color
different configurations of the causal variables that Processing Present] exist in corporate logo Then
produce the outcome variable. And the goal of [Creative-innovative Image] exists. Extracted
logical minimization is to represent the data in a sets indicated that in order to impart the image of
rationally constructed shorthand manner as the "Familiar- Favorable" in logo design it is important
Table 6 below demonstrates [22,23] The sets that the language character used in logo attributes
shown in this table were minimized by contraction to be known by the consumer. Moreover, the
technique for interpretation by the following image of Creative-Innovative correlates directly
complementary rules were applied based on with [Logo Entirety] and [Logotype], and not
decision class weights (0.9), (0.6), and (0.444) necessarily with [Logo Mark], it reveals that
respectively : 1) If [known language] exists in "Creative-Innovative" image in logo design can be
corporate logo Then ["Familiar- Favorable" Image] imparted by the combination of logo attributes used
exists, 2) If [Solid sense present] And [known in logo design not by concentrate on a specific
language] exist in corporate logo Then attribute such as active sense in Coca Cola or
[Traditional-Consistent Image] exists, and 3) If thickness sense in FedEx logo.
[Logo Entirety Active]
Table 6. Contraction
2 0.6 0.6
Traditional
Consistent
0.6 0.6 0.6 0.6
6 CONCLUSION & FUTURE WORK familiarity in logo image with the familiarity of
language used in logo design to be recognized by
Inspection of the factor loadings in both surveys Japanese, it suggests that the familiarization of
prioritized three influential images significantly Arabic language to Japan consumers is important
and substantially different from zero value to promotion Arab corporate and products in Japan
Familiar-Favorable, Creative-Innovative, and marketplace. But in spite of the unfamiliarity in
Traditional-Consistent. This clarifies that logo Arabic language, Arabic alphabetic logotypes seem
design characteristics are captured by the same to impart a strong image of creativity for Japanese
factorial structures and also strongly outperforms people, perhaps because of hidden similarity
simultaneous basics for Arabic logotypes design between Arabic alphabets and Japanese characters.
industry in commercial sectors with three Future research can extend our findings in several
dimensions: 1) Familiarity, 2) Creativity, and 3) meaningful ways for upcoming 2020 Olympic.
Traditionalism. Respectively, the analyses of Most notably a cross-cultural research between
clusters as well as average scores comparison in Arab and Japan investigating Arabic and Japanese
survey I and Survey II confirmed that English language calligraphy psychological influence on
alphabetic logotypes are familiar to Japanese ones personality traits.
people than Arabic alphabetic logotypes most
likely due to the influence of American culture.
Also since rough sets analyses linked between
8
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 1-9
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
9
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 10-17
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
10
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 10-17
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
Holton, D.R.W., Younas, M., and Awan, I.U. In order to control accesses, NAP-Web uses
developed and evaluated the system that was Run_Ready (RR), Wait queue (WA) and
assumed to be used in e-commerce area[3]. They Re_Access queue (RA) and Next_Wait table.
contended that the requests received from a They are shown in left part of Figure 1. RR
Web portal should generally get higher priority stores currently processing accesses. Thus the
as such requests are more likely to lead to size of RR decides the maximum processing
purchases. Received access is stored in the number of accesses. By limiting the size of RR,
queue corresponding to type of it. If priority of Web server load can be suppressed. WA stores
incoming access is higher than that of current accesses waiting for processing by FIFO manner.
processed access then it is suspended and Next_Wait keeps ticket information to identify
incoming access is processed. access whether it is re-accessed or not. RA stores
re-accesses by FIFO manner. When RR is not
Elnikety, S., Nahum, E., Tracey, J., and full, the incoming access is entered in RR. When
Zwaenepoel, W. determined the priority of a RR is full and all accesses in WA and RA are
process according to the size of job[4]. By expected to finish within upper limit time, it is
lowering the priority of a big job, it raises the queued in WA. In our experiments the upper
overall responsiveness. Cherkasova, L. and limit time is set 4 seconds. Otherwise, the ticket
Phaal, P. introduced a session based admission is created, recorded on Next_Wait and returned
control to ensure that longer sessions can be the client which sends the access. Thus NAP-
completed[5]. If a server is functioning near its Web does not pile up accesses waiting for TCP
capacity, a new session will be rejected. These connection, and new connection request can be
researches emphasize to prevent overloading and established immediately. If re-access reaches at
maintain total throughput at peak rather than to the notified time, it is queued in RA. When RR
ensure the response time of each classes. goes to not full, the head of WA goes to RR if
Menasc, D.A., Fonseca, R., Almeida, V.A.F.,
and Mendes, M.A. present a family of priority-
based resource management policies for e-
commerce servers[6]. This aims at optimizing
revenue/sec instead of focusing on conventional
performance metrics. Their target is
improvement of responsiveness to clients and
single Web server is assumed. In our research,
responsiveness is improved by using multiple
cache servers, and the quality of resource served
by them is raised by priority processing.
Figure 1. Priority access mechanism based on NAP-Web
11
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 10-17
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
WA is not empty, and the head of RA goes to it could adjust the quota of computer resource to
WA if RA is not empty. process priority access by changing the ratio of
the size of RR and PRR. The throughputs of both
We now describe Priority Access Mechanism groups were measured with several combination
shown in right part of Figure 1. Accesses are of the size of RR and PRR.
classified into two groups. One is priority access
group handled by Priority Access Mechanism. 4.1.1 Experiment Description
The other is normal access group handled by
original NAP-Web. After this we call original Experiment environment is shown in Figure 2.
NAP-Web Normal Access Mechanism. Each client issues the designated simultaneous
Pri_Run_Ready (PRR) stores currently accesses severally to a bulletin board system
processing priority accesses. Pri_Wait queue (ASKA BBS [9]) on origin server. Each client
(PWA) stores priority accesses waiting for generates designated number of threads. Each
processing. By limiting the number of thread accesses repeatedly to origin server
simultaneous priority accesses under the size of simultaneously. The upper limit of the response
PWA, priority accesses are always processing or time of the NAP-Web is set to 4 seconds. The
waiting with keeping TCP connection. Normal combination of size of RR and PRR used in the
Access Mechanism and Priority Access experiments is shown in Table 1. (n, m) shows
Mechanism are placed in parallel. The number of (the size of PRR, the size of RR). The total
current processing accesses is sum of the number number of threads on normal clients is 3,000
of accesses in RR and PRR. which is enough number to make origin server
stay in overload. The number of threads on
We divert queues of NAP-Web except priority client is 5 which is enough number to fill
Next_Wait and RA to Priority Access PRR. Normal clients start access at the
Mechanism and add selector. By changing the beginning of experiments and priority client
maximum number of access in RR and PRR, starts access at after 200 seconds later. The
system can control computer resources for reason to delay starting time of priority client is
processing priority access. In our system, source that re-access time prediction is disordered just
IP address is used to identify priority access. after starting normal access.
Selector checks source IP address of reached
access. If it is one of one of IP addresses of
cache servers, the selector passes the access to
Priority Access Mechanism. Otherwise, it passes
the access to Normal Access Mechanism. In
experiments in Section 4, IP addresses of priority
client and cache server are hard coded.
12
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 10-17
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
4.1.2 Results of Experiments (1, 10) (1, 20) (2, 10) (2, 20)
Throughput
Graphs are separated into 2 groups in all
parameters. This is because there are 2 kinds of 100
accesses. One succeeds at the first time access,
50
and the other requires re-access. Response time
required re-access is sum of the processing time 0
in the first access, the waiting time and the 100 200 300 400 500 600 700 800
processing time in the re-access. The lower Time (s)
group includes accesses that succeed at the first
time, and the upper group includes accesses that Figure 4. THp, THn and THt
succeed by re-access. Response time in lower Table 2. The Percentage of Throughput
group is about 4 seconds which is almost equal Percentage of PRR
Average percentage of
to upper limit stay in WA and RA. This shows in sum of PRR and
THp in THt
that maximum staying time in WA and RA is not RR
so influenced by priority access. Response time (1, 10) 9.1 21.4
(1, 20) 4.8 22.0
without priority access (0, 20) in upper group is (2, 10) 16.7 31.3
about 20 seconds in all time. However response (2, 20) 9.1 27.8
time after starting priority access (200 seconds
point) in upper group lengthens and is disordered. equal to the percentage of PRR. The percentage
The response time with high percentage of PRR of THp is, however, higher 12.3% to 18.7% than
is longer than that of low percentage of PRR. that of PRR. Moreover the percentage of THp
with (1, 10) is different from that of THp with
Priority access throughput (THp), normal access
(2, 20). On the other hand, the percentage of
throughput (THn) and total throughput (THt) are
shown in Figure 4. They are placed lower, center THp with (2, 20) is higher than that of THp with
and upper location, respectively. THt is sum of (1, 20), and the percentage of THp with (2, 10) is
THp and THn. THt is about 160 in all also higher than that of THp with (1, 10).
combinations even if the sizes of the RR and Magnitude relation between percentage of THp
PRR are changed. It shows that THt is not and percentage of PRR with same size of RR is
influenced by the size of them. Table 2 shows preserved. This result shows that it is possible to
the percentage of PRR in sum of PRR and RR adjust the quota of computer resource to process
and average percentage of THp in THt for 600 priority access but it is insufficient. Improvement
seconds (between 200 to 800 seconds). We of adjustment precision is future work.
expected that the percentage of THp was roughly
4.2 The Effect on Response Time to Clients
Through Cache Server
(0, 20) (1, 10) (1, 20)
(2, 10) (2, 20) The purpose of this experiment is to confirm that
40 Priority Access Mechanism can reduce response
(s)
time to clients through cache server. As shown in
30 Figure 5, we use origin server with our
Time mechanism, cache server, direct clients accessing
20
origin server, and indirect client accessing cache
10
server. While origin server is overloaded by
Response numerical simultaneous accesses from direct
0 clients, indirect client accesses to cache server.
100 200 300 400 500 600 700 800 The expiration date of cache is set a few seconds
Time (s) for caching dynamic contents. The response time
Figure 3. Response time of normal access
to direct and indirect clients is measured.
13
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 10-17
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
14
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 10-17
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
In the following subsections, target recognition Authentication CGI is the interface of Priority
method, design and function evaluation of Access Management Mechanism, and used for
Dynamic Registration Method are described. registers cookie. It examines whether user who
want to register his host is legal or not, and then
5.1 Consideration of Target Recognition requests priority key to API and returns priority
Method cookie including priority key as shown in Figure
10. Priority key has an expiration date, and it is
IP address and cookie can be used to recognize deleted automatically if there is no access for a
whether an access is priority access or not. In our certain period of time.
implementation of Priority Access Mechanism
used in Section 4, IP addresses of priority client API has interfaces that generate, register and
and cache server are stored in IP address array return priority key, and register and delete
15
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 10-17
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
16
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 10-17
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
renewed in a short time by Priority Access commerce Servers, ACM SIGMETRICS Performance
Mechanism and then short term cache for Evaluation Review, Vol.27, Issue 4, pp.27-35, 2000.
dynamic contents could be possible. And the 7. Kaji, T. and Saisho, K., Proposal of Web System
NAP-Web to Promise Next Access, Proceedings of
mechanism can adjust the quota of the computer International Conference on Instrumentation, Control
resource to handle access from cache server. and Information Technology (SICE Annual
However adjustment precision is insufficient. Conference 2007), pp.959-964, 2007.
We also confirm that Priority Access 8. https://httpd.apache.org/
Management Mechanism can register priority 9. http://www.kent-Web.com/bbs/aska.html
hosts dynamically and accesses issued from
registered hosts are processed by Priority Access
Mechanism.
REFERENCES
17
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 18-25
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
18
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 18-25
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
Figure 1. Experiment Environment The typing test was performed untimed until the
participants become used to the task. The
number of characters used were one-third that of
the main experiment. The typing system was the
same as that in the main experiment.
19
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 18-25
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
20
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 18-25
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
21
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 18-25
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
P1 P2 P3 P4 P5 P6 P7 P8
Active
- 0 0 - - - - 0
Anger
++ 0 + + + ++ 0 +
Comfort
0 ++ - - ++ ++ ++ +
Relaxed
++ ++ + + + + - ++
Strained
0 + + + 0 0 0 +
22
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 18-25
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
As shown, the scores of many participants were second experiment was greater than the number
clearly improved in the items of of input errors in the first experiment.
Anxiety/Uneasiness and Comfort.
Six participants had clearly positive 3.2 Physiological responses to Wrist Warming
psychological effects of Anxiety / Uneasiness
as shown in Figure 5. Moreover, four To confirm change of TEMP, GSR, and HR, the
participants had clearly positive Comfort following data were extracted from BioTrace+:
effects as Figure 6 shows.
PRE-EX: Last 30 seconds before the first
Improved calm period elapsed.
Positive Effects EX-FIN: Last 30 seconds before the typing
test ended.
Normal Warm
23
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 18-25
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
P1 P2 P3 P4 P5 P6 P7 P8 P1 P2 P3 P4 P5 P6 P7 P8
Normal Normal
Warm Warm
From the result, the finger skin temperature From the result, it seems possible that increased
(TEMP) increased for many participants via finger skin temperature by wrist warming
wrist warming. Therefore, it is possible that they increases GSR. Therefore, it seems possible that
experienced stress relief by wrist warming. the sensors could not accurately detect the stress
Moreover, there is a high possibility that they relieving effects of wrist warming.
obtained these stress relieving effects by the
warming of their fingers whose temperatures had 3.2.3 Influence on HR
decreased because of the mental workload.
Figure 9 shows EX-FIN and REST-FIN for HR,
3.2.2 Influence on GSR and Table 7 shows the effect obtained from the
reduction rate between EX-FIN and REST-FIN.
Wrist warming decreased GSR for some In this result, no significant difference was
participants. However, many of them had the detected between the warm and normal armrests.
result of their GSR decreased in normal
temperature armrest than in wrist warming.
Normal Warm
Figure 8 shows amount of change from PRE-EX,
and Table 6 shows the effect obtained from the
difference between EX-FIN and REST-FIN.
Normal Warm
Figure 9. Changes in HR
Table 7. Effects of HR
Figure 8. Changes of GSR
P1 P2 P3 P4 P5 P6 P7 P8
Normal
*
Warm
*
*1 participant data was excluded cause by sensor measurement error
24
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 18-25
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
In this experiment, we had measured ECG data 3. The Ministry of Health, Labour and Welfare,
Guidelines for Industrial Health Controls of VDT
from both arms of each participant. Operations, Japan Kihatsu No. 0405001, April 5,
1 participant could not detect normal heart rate 2002.
4. M. Suzuki, N. Hirano, H. Terashita, Y. Oda and
because of interference by muscular motion. A. Yagi, On the items and usage of a questionnaire
Moreover, it seemed possible that other for evaluation of stress caused by phasic task
demand, Japan Ergonomics Society vol 35(4),
participants HR data had some influence due to pp.259-270, 1999.
same reason. 5. A. Sato, and A. Yasuda, Development of the
Japanese version of Positive and Negative Affect
Schedule (PANAS) scales, Japan Society of
4 CONCLUSION Personality Psychology vol. 9(2), pp.138-139, 2001-
03-30.
6. K. Hashimoto, M.Murakami, Reliability and validity
In this study, we researched the psychological of the reversed Mood Check List-short form 2 (MCL-
and physiological responses to wrists warming S.2) measuring the positive mood state following
after mental workload from a timed typing test. exercise, Jounal of Health Science, Kyusyu
Psychological responses were evaluated with a University, Vol.33, pp.21-26,2011
21 questionnaire items with 7 mood state
categories.
Physiological responses were evaluated with
changes in the heart rate, skin temperature and
galvanic skin temperature.
We conducted an experiment involving 8 healthy
participants.
REFERENCES
1. K. Kato, Properties of Pleasure Using Hot Back-
Compresses : Changes in Pleasure-Displeasure and
Autonomic Nerve Activity after the Uchida-
Kraepelin Stress Test, Japanese Journal of Nursing
Art and Science Vol. 9, No. 2, pp 413 2010
2. K. Kaneko, Y. Ogata, H. Kumagai, M. Yamamoto,
Physiological Effects of Hot Compresses to Foot
with Steamed Towel : Comparison with Footbath
Uses on Cardiovascular Dynamics and Autonomic
Nervous Functions, Japanese Journal of Nursing
Art and Science Vol. 11, No. 2, pp4-11, 2012
25
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 26-34
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
Table 2. The drivers types in conjunction with moving types.
fast moving sure moving
f-nb-nf s-nb-nf
f-nb-f s-nb-f
f-b-nf s-b-nf
f-b-nf-a s-b-nf-a
f-b-f s-bf
f-b-f-a s-b-f-a
Figure 7. Six cases of driving in our traffic model. The speed of each car at every time steps is stochas-
We consider six car driving types based on the tic, and we assume that it, the speed of all the cars,
driver effects shown in Table 1, and assign the sym- has the same probability density distribution f (x) in
bols nb-nf,.., and b-f-a to their driving types. We also the cases without driver effects. But, the form of
assign the symbols to their drivers types in conjunc- f (x) varies diverse if there are driver effects. We
tion with moving types as shown in Table 2. can consider that a traffic is one of stochastic moving
We can consider several cases to run a traffic as multi-agents with exclusive resource sharing: each
shown in Figure 7. The outline of our algorithm is cell is occupied by a car at most, and a traffic flow
given as the following: Suppose a car is running on a is a behavior of the multi-agents whose agents are
traffic. arranged on a straight line consisting of traffic cells.
Our driver effects provide the interaction between two
1. Minimum boundary effect: If the car runs agents, but no direct interactions among more than
in minimum boundary effect, it may be either three agents.
shorten or not shorten. If it is true and the car In the following section, our experiments show how
speed is not zero, the car moves to the next cell. the driver effects are related to the traffic flow in two
Otherwise, it will stay there. moving types.
29
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 26-34
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
3 EXPERIMENTS most even, i.e. the distribution of it is almost flat.
Then, the traffic flow versus car density relationship
has moved to the right upside. But, the traffic flow
We consider six car driving types based on the driver and car density are low in both in b-f. The similar
reactions in Section 2, shows them in Table 1, and trend is seen in Figure 10. When the speed distri-
assign the symbols nb-nf,.., and b-f-a to their driving bution is sharpened more slightly (Figure 11), b-nf-a
types. turns to left upside, and others are moving toward the
We prepare a one-way traffic consisting cells, and upside. Furthermore, a little more sharped case of the
the traffic is one lane. Every car moves on the traffic speed distribution is shown in Figure 12. The Figure
cells from left to right shown in Figure 1. The num- 13 shows the most sharpened case. There, the traffic
ber of the traffic cells is 2, 000. The moving type is flow and car density are still low in both in b-f. b-nf-
fast moving, so we perform the updates of the traffic a, b-nf and nb-nf, and they are coming together to the
cells from right to left, reverse, at each moving step, middle of upside. On the other hand, b-f-a moved to
and provide 20, 000 cars. Their cars run 20, 000 steps left upside. We examined the difference between n-
for each exam in our experiments. At each updating nf-a and b-f-a in Figure 14. The more detail is shown
step, a car enters into the traffic only if the leftmost in Figure 15. The difference between two is only fluc-
traffic cell is empty. Otherwise, the car has to wait tuation effects. Both are almost the same traffic flow,
to enter into the traffic until left most cell becomes but quit different in car density. So, we gradually
empty. This means that we examine the maximum change the event probabilities of the fluctuation ef-
traffic flow for every driving type in fast moving. fects between b-nf-a and b-f-a. res-f-00-16-00 repre-
Each car speed randomly changes at every moving sents that there are no fluctuation effects. res-f-01-14-
step. We use Mersenne twister as a random number 01 represents that it increases the speed 1 in average
generator. Then, each car speed v is either 0, 1, 2, 3 with probability 16 1
, and it decreases the speed 1 in
or 4 in the cases without driver effects, and the distri- 1
average with probability 16 . Otherwise, there are no
bution of their speed is given by 0, 1/16, 14/16, 1/16
fluctuations. Similarly, res-f-08-00-08 represents that
and 0 in this examination. So the average value is 2 if 8
it increases the speed 1 in average with probability 16 ,
there are no effects.
and it decreases the speed 1 in average with probabil-
The boundary effects are implemented as every car 8
ity 16 . Then, the traffic flow of res-f-03-10-03 and
speed v either increase 1 in average if the behind cell
res-f-04-08-04 has remained in a small change, while
is occupied and two more cells in ahead are unoccu-
their car densities are quit distinct. It works like a
pied, or decrease 1 in average if v more cells in ahead
switch in electrical circuits. Figure 16 shows the case
are unoccupied and the car approaches the behind of
that every car speed is 2, while it allows the mini-
a car in ahead with lower speed. Shortening one cell
mum boundary effects. By changing the distribution
in the minimal boundary effect (Figure 5) is consid-
of fluctuation effects, we can see that the traffic flow
ered if there only exists an empty cell in front and the
keeps higher while the car densities are changed dras-
speed is greater than or equal to 1. With even proba-
tically.
bility, the fluctuation effects either increase the speed
Figure 21 shows how boundary effects are related
1 or decrease the speed 1 in average.
to fluctuation effects, so that both boundary and fluc-
We examine the car density and the traffic flow by
tuation effects are of essential to achieve efficient uti-
varying the distributions of the speed for every exam
lization of traffic resources, where b, f and d repre-
(Table 3). Then, we observe the car density on the
sent the probabilities of boundary effect, the std. de-
cells between 501th and 1,500th, and the traffic flow
viation of fluctuation effect and car density on cells,
is the number of cars to pass on the 2,000th cell be-
respectively. Our results show that we feel stress in
tween 10,001th and 20,000th steps. We repeat 10
driving and take some actions as driver reactions so
times for each driving type.
that the car density in traffic flow is improved drasti-
We obtained the traffic flow versus car density rela-
cally. The figure in Figure 22 shows the number of
tionships shown in Figure 8-13. The probability dis-
traffic flow varying the probability of boundary and
tribution of each car speed is shown in Table 3. In
fluctuation effects in two dimensional space, where v
Figure 8, the car speed splits into either low or high.
represents the number of cars to pass. Then, the car
Then, the traffic flow and car density are low in both
density becomes higher in fast moving and minimum
in the car driving type b-f. If we disallow fluctua-
boundary.
tion effects and allow to shorten one cell in minimum
The figures in Figure 17 and 18 present the car den-
boundary, that is b-nf-a, the traffic flow and car den-
sity and the number of traffic flow. Then, the mixture
sity become high in both.
of fast moving and sure moving with probability 0.5,
In Figure 9, the probabilities of each speed are al-
30
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 26-34
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
Table 3. The experimental table: the average speed is 2 for each, pends on the driver effects, and we can see the phase
and their distributions are listed in below. transitions under some traffic conditions.
speed 0 1 2 3 4
5 3 3 5 REFERENCES
Figure 8 16 16 0 16 16
3 3 4 3 3
Figure 9 16 16 16 16 16 1. Bando M., Hasebe K., Nakanishi k., Nakayama A., Shi-
2 3 6 3 2
Figure 10 16 16 16 16 16 bata A., Sugiyama Y.: Phenomenological Study of Dynam-
1 2 10 2 1 ical Model of Traffic Flow, J. Phys. I France, vol. 5, 1389
Figure 11 16 16 16 16 16
1 1 12 1 1 1399(1995),
Figure 12 16 16 16 16 16
1 14 1
Figure 13 0 16 16 16 0 2. Clarridge A., Salomaa K., Analysis of a Cellular Automaton
Model for Car Traffic with a Slow-to-Stop Rule, Lecture
Notes in Computer Science, vol. 5642, 4453(2009).
3. Derrida B., Evans M.R., Hakim V., Pasquie V.: Exact solu-
tion of a 1D asymmetric exclusion model using a matrix for-
mulation, J. Phys. A: Math. Gen., 278287, vol. 26(1993).
4. Nagel K., Schreckenberg M.: A cellular automaton model
for freeway traffic, J. Phys. I France, vol. 2, 2221
2229(1992).
5. Nieuwstadt F.T.M., Steketee J.A.: Selected Papers of J.M.
Burgers, Springer (1995).
6. Schadschneider A., Schreckenberg M.: Cellular automaton
models and traffic flow, J. Phys. A: Math Gen., vol. 26,
d L679L683(1993).
7. Shioya I.: Resource Utilization of Stochastic Mobile Multi-
Agents, IEEE PACRIM, 137141(2013).
Figure 8. Traffic flow versus car density relationships in car driv-
8. Shioya I.: An Accelerated Resource Utilization in Au-
ing types: the probabilities for the speed 0, 1, 2, 3 and 4 corre-
5 3 3 5 tonomous Stochastic Mobile Multi-Agents, International
spond to 16 , 16 , 0, 16 and 16 , respectively.
Journal of Innovation in the Digital Economy, vol. 5, 1
and it allows minimum boundary effect. There, we 14(2014).
varied the probabilities of boundary effects and fluc- 9. Shioya I.: Resource Acceleration of Autonomous Stochas-
tuation effects in two dimensional space. Then, the tic Moving Multi-Agents with Boundary Effects, The Inter-
car density is almost all to keep the constant value national Conference on Artificial Intelligence and Pattern
0.4. Also, the number of traffic flow keep the con- Recognition, 134142(2014).
stant value 6500. 10. Shioya I.: A traffic flow analysis from psychological as-
The figures in Figure 19 and 20 show the car den- pects, ICDIPC-IEEE, 712(2015).
sity and theh number of traffic flow in sure moving,
and it allows minimum boundary with probability 0.5.
Then, the highest density can be seen in no boundary
and allow fluctuation, i.e. s-nb-f-a(0.5).
Our result shows that, in sure moving, the fluc-
tuation effect is useful, but not boundary effect. In
fast moving, the fluctuation effect is useful to achieve
high traffic resource utilization, and it also related to
4 CONCLUSIONS
d
d
Figure 10. Traffic flow versus car density relationships in car Figure 13. Traffic flow versus car density relationships among
driving types: the probabilities for the speed 0, 1, 2, 3 and 4 car driving types: the probabilities for the speed 0, 1, 2, 3 and 4
1 14 1
2
correspond to 16 3
, 16 6
, 16 3
, 16 2
and 16 , respectively. correspond to 0, 16 , 16 , 16 and 0, respectively.
d
d
d
d
32
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 26-34
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
0-1-14-1-0
0.8
0.7
0.6
1 0.5
0.8
0.6
d
0.4
0.2
0
1.2
1
0.8
0 0.6
0.2 0.4 f
0.4
b 0.6 0.2
0.8
1 0
d
Figure 19. The car density in sure moving, and minimum bound-
ary with probability 0.5: the probabilities for the speed 0, 1, 2, 3
1 14 1
Figure 16. Traffic flow versus car density relationship between and 4 correspond to 0, 16 , 16 , 16 and 0, respectively.
b-nf-a and b-f-a by changing the distribution of fluctuation ef-
fects: the probabilities for the speed 0, 1, 2, 3 and 4 correspond
to 0, 0, 1, 0 and 0, respectively.
0-1-14-1-0
9e+003
8e+003
0-1-14-1-0
0.4 10000
0.3 9000
8000
1 v 7000
6000
0.8 5000
0.6 4000
d 3000
0.4
1.2
0.2 1
0 0.8
0 0.6
1.2 0.2 0.4 f
0.4
1 b 0.6 0.2
0.8
0
0.8 1 0
0.2 0.6
0.4 0.4 f
b 0.6 0.2
0.8
1 0
Figure 20. The traffic flow in sure moving, and minimum bound-
ary with probability 0.5: the probabilities for the speed 0, 1, 2, 3
Figure 17. The car density in the mixture of fast moving and sure 1 14 1
and 4 correspond to 0, 16 , 16 , 16 and 0, respectively.
moving, and allow minimum boundary with probability 0.5: the
1
probabilities for the speed 0, 1, 2, 3 and 4 correspond to 0, 16 ,
14 1
,
16 16 and 0, respectively.
0-1-14-1-0
0.9
0.8
0-1-14-1-0 0.7
6e+003 1 0.6
5e+003 0.8 0.5
10000 0.6
d
9000 0.4
8000 0.2
v 7000
6000 0
5000 1.2
4000 1
3000 0.8
0 0.6
1.2 0.2 0.4 f
1 0.4
b 0.6 0.2
0.8 0.8
0 0.6 1 0
0.2 0.4 f
0.4
b 0.6 0.2
0.8
1 0
33
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 26-34
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
0-1-14-1-0
10000
9000
8000
v 7000
6000
5000
4000
3000
1.2
1
0.8
0 0.6
0.2 0.4 f
0.4
b 0.6 0.2
0.8
1 0
Figure 22. The traffic flow in fast moving and minimum bound-
ary: the probabilities for the speed 0, 1, 2, 3 and 4 correspond to
1 14 1
0, 16 , 16 , 16 and 0, respectively.
34
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 35-41
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
35
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 35-41
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
36
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 35-41
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
In order to find enough improvement goal, an level is lower than zero, it means the supplier can-
equation for calculating coefficient of prospective not become efficient by decreasing the changeable
level of changeable factors is proposed. Manufac- inputs.
turers can make decision for improving or chang-
ing suppliers based on values of coefficient of S ( ni=1 i Xi ) m u X u
S = nm c ci=1 i i (1)
prospective level. i=1 i Xi
The steps to find out weak links in supply chains
are described as follows: Where
37
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 35-41
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
Figure 2. Structure of two stage chain with two suppliers and one manufacturer.
The results indicate that there are seven inefficient 4.3 Calculate coefficient of prospective level
chains, i.e. chain 5, 6, 7, 8, 11, 12 and 13. Consid- of input (Step3)
ering efficiency score of supplier1 , there are seven
Coefficient of prospective level of changeable in-
efficient chains, i.e. chain 1, 3, 4, 7, 9, 10 and
put is calculated with the equation (1). Under the
15. While only four chains that consist of ef-
condition of this case, equation for calculating co-
ficient supplier2 , i.e. chain 4, 9,14 and 15. In
efficient of prospective level of changeable inputs
other words, only chain 4, 9 and 10 consist of
is shown as follows:
efficient suppliers, and it means only these three
chain are good enough in the current stage. Thus, S (1 X1 + 2 X2 + 3 X3 ) (1 X1 + 2 X2 )
to improve the chains performance these ineffi- S =
3 X3
cient suppliers should be improved. (2)
Where
38
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 35-41
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
Table 2. Data sets of the representative chains in the frozen shrimp industry.
Chain X1 X2 X3 Y1 Y2 Z1 Z2
1 s1 8.21 271.30 42.50 97.14 96.18 0.36 98.53
s2 7.49 306.90 42.50 96.63 93.23
2 s1 12.05 306.90 42.50 96.93 95.27 0.66 96.92
s2 4.46 271.30 42.50 96.49 92.23
3 s1 12.06 282.80 40.00 97.12 96.03 1.23 98.62
s2 6.58 306.90 40.00 95.00 91.88
4 s1 13.40 271.30 45.00 97.62 94.24 2.33 97.65
s2 7.67 271.30 43.00 96.00 91.09
5 s1 14.11 373.80 45.00 96.97 95.20 2.72 97.08
s2 4.97 373.80 47.00 96.56 91.83
6 s1 16.51 306.90 47.00 98.22 95.57 3.41 97.18
s2 5.20 282.80 42.50 96.40 91.68
7 s1 20.37 271.30 42.50 97.69 93.97 3.46 97.74
s2 6.00 373.80 45.00 96.62 91.59
8 s1 21.03 373.80 42.50 97.39 94.90 3.97 97.31
s2 5.24 282.80 42.50 96.04 91.79
9 s1 24.42 347.30 32.50 97.08 94.39 3.97 98.42
s2 4.89 271.30 35.00 96.39 92.29
10 s1 27.87 306.90 35.00 98.04 95.58 4.31 97.45
s2 4.68 347.30 40.00 96.91 91.85
11 s1 29.04 309.50 45.00 97.14 95.96 4.71 97.68
s2 5.45 306.90 45.00 96.05 91.53
12 s1 29.85 282.80 43.00 97.17 94.71 9.79 98.06
s2 4.03 282.80 47.00 95.89 92.34
13 s1 36.59 373.80 42.50 96.96 95.47 10.08 97.61
s2 4.00 373.80 42.50 95.57 92.44
14 s1 37.41 282.80 40.00 96.77 95.55 16.86 97.35
s2 3.95 309.50 35.00 96.30 91.42
15 s1 38.60 271.30 42.50 97.69 94.41 18.43 98.12
s2 3.51 271.30 40.00 96.99 92.45
39
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 35-41
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
Table 3. Efficiency results of the representative chains in Table 4. Coefficient of prospective level of logistics costs
frozen vegetable industry. Asis means efficiency score of of frozen shrimp industry when considering input-oriented
chains. S1 means efficiency score of supplier1 . S2 means perspective. means coefficient of prospective level of sup-
efficiency score of supplier2 pliers
ering output-oriented strategy or fundamental in- 6. Chaowarat, W., Shi, J.: Efficiency measurement of
vestment. Such cases also be dealt by modifying supply chain in the export frozen vegetable industry:
the proposed model as output-oriented strategy. A network DEA approach: International Journal of
Materials, Mechanics and Manufacturing 1(1), 8891
(2013).
ACKNOWLEDGMENT
This work was supported by Grants-in-Aid 7. Tseng, M.L., Chiang, J.H., Lan, L.W.:Selection of op-
timal supplier in supply chain management strategy
for Scientific Research from Japan Society for with analytic network process and choquet integral:
the Promotion of Science (NO. 15K00486, Computers and Industrial Engineering 57(1), 330340
26350013.) (2009).
40
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 35-41
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
15. Chen, C., Yan, H.: Network DEA model for supply
chain performance evaluation: European Journal of
Operational Research 213(1), 147155 (2011).
41
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 42-48
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
42
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 42-48
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
43
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 42-48
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
frames will be delivered across the wireless new retransmission and presenting the same
channel. power. Note that, the retransmitted packet is
similar to the discarded one. The retransmission
3.1. MCS scheme principle operation is repeated until a successful reception
of the erroneous packet. Beyond that thanks to
Considering the physical layer, the convenient truncated HARQ protocol features if the
MCS mechanism is chosen according to the repeated retransmission of the erroneous packet
channel condition where the impact of reached the maximum resending number K, the
estimations errors is not considered in our packet will be rejected and outage event
analysis. Typically, MCS selection is based on happens. Error detection is assumed perfectly
dividing the overall SNR range into I+1 small performed by the CRC pattern. In addition, the
intervals denoted by [ ) wherein i = same power is equally distributed at each
0,1,I+1, in which following the convention retransmission attempt. Following this modeling
= 0 and = . It should be noticed that process, if we assume a fixed length of the
MCSi presents a rate Ri. Specifically, the MCS transmitted packet and taking into account the
scheme is selected according to the retransmission attempts denoted by k, the
instantaneous CSI which is supposed available at received signal could be expressed by:
both the transmitter and the receiver. In addition,
the returned feedback is assumed without latency = + (1)
and without estimations error. Indeed, the
receiver notifies the transmitter about the Above denotes the transmitted symbol at the
th
suitable MCS according to the received SNR l subcarrier stands for the channel
which then justifies the suitable transmitted th
coefficient at the k retransmission attempt.
power which improves the EE. corresponds to the noise.
3.2 Joint Coupling between MCS scheme and 4 ENERGY EFFICIENCY OPTIMIZATION
HARQ mechanism
4.1 Energy consumption model
In this subsection we explain the association Energy expenditure measuring depends on the
between MCS scheme and SR HARQ type I MB-UWB cross-layer design in which hardware
retransmission protocol. Facing to fading structure should be considered. Especially, the
problems HARQ retransmission protocol is energy consumption model is depicted in Fig.1.
required due to the reliability that they could
ensure. Otherwise, MCS scheme is widely
employed for throughput maximization for
ARQ
instance in [16]. Motivated by the reliability Controller
requirement we have applied SR HARQ type I
retransmission protocol of the data link layer. Input Adaptive
Buffer Coding IFFT RF
According to fading problems, we have Packets modulator
Retransmission Request
44
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 42-48
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
4.2 Energy expenditure per bit minimization rate in net shape in which accounts the bit
information per packet length. accounts the
EE examination demands several features Peak to Average Power Ratio
investigation like PER and MCS rate. For this, (PAPR). Moreover, for sake of simplicity we
the aim of the optimization approach is to adopt the PER formula derived in [5] as:
investigate MCS rate impacts on energy
expenditure per bit minimization under the
respect of PER. Interestingly, total energy
consumption per bit diminishing provides EE
improvement. Thus, the optimization issue of (6)
determining the optimal MCS rate enabling
energy expenditure diminishing can be Note that and could be identified through
represented by: curve-fitting method where relies to SNR
threshold level such that tabulated in
(2) Table I. In other side, corresponds to the
smallest Signal to Noise Ratio (SNR) value
identifying the PER corresponding to a specific
Above Ea denotes total energy consumption per MCS rate.
bit. R (bit/sym) stands for MCS rate. Henceforth
refers to PER. Correspondingly, the Due to Ea convexity ( , the
achievable MCS rate selection is denoted by Ri optimization challenge could be solved by
which fundamentally depends on the CSI. Thus,
setting which implies
after the repeated retransmission attempts the
overall energy consumed per bit could be
quantized by: (7)
MCS Data
(3) Coding Mapping
level rate ai bi i (dB)
rate (r) (M)
(Mb/s)
being: 1 53.3 1/3 4 0.69 0.212 -2.23
Gl stands for the gain factor at d = 1m. Ml 10 960 1/2 16 0.67 2.32 1.24
compensates the variability due to hardware
process. stands for the noise density.
corresponds to the effects of baseband By this way we write:
processing. L accounts the number of uncoded
bit. L0 denotes the overhead bits. stands to the
filter roll-off factor. is the drain efficiency of
the amplifier. denotes the information
45
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 42-48
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
5 PERFORMANCE ANALYSIS
Numerical results are performed into the MB-
UWB systems following the ECMA-368
(10) standard according to the first three sub-bands.
Indeed simulations results are carried out over
Secondly, we denote respectively by and CM1 IEEE 802.15.3a channel model. In addition
the values of SNR and R related to the for simplicity reasons we adopt the features of
previous MCS scheme where we have: the energy consumption model defined in [18]
such that Ml = 40 dB, Gl = 30 dB, = 3.5, N0/2
= -174 dBm/Hz. In addition, B = 1024 bytes and
K1 = , B= , = the other parameters are defined as [12] such that
(11) kb/s and B0= 48. Additionally, we
exploit the MCS parameters reported in Table I
Which leads that (10) could be expressed as: with a CRC-32 component. The maximum
number of retransmission attempts is K = 2 with
a target PER = 10-5. Hence, the proposed
(12) algorithm for the most energy efficient MCS
guaranteeing energy per bit minimization is
validated in Fig.2 in which we compare the
Yields: results of the energy consumption per bit relied
to the optimal MCS rate with other MCS
schemes when varies between 10-1 and 10-6.
Otherwise, the energy per bit is simulated for
G(R)= (13)
various combination of MCS with HARQ type I.
46
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 42-48
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
40
significant minimization of energy per bit which
Proposed Method (Optimum MCS)
MCS 1 is the requirement of EE.
35
MCS 5
MCS 10
30 6 CONCLUSION
Energy per bit (dBmJ)
25
25
computer results which reveal the effectiveness
20
of the proposed algorithm in terms of energy
15 expenditure diminishing. Further contributions
10
should consider the delay towards the returned
request. In addition, estimation's errors should be
5
included in the optimization issue.
0
-6 -5 -4 -3 -2 -1
10 10 10 10 10 10
0
REFERENCES
Figure 3. Energy consumption per bit comparison results 1. FCC Document 00-163, Revision of Part 15 of the
between our solution and other methods. Commissions Rules Regarding Ultra-Wideband
Transmission Systems., (2002).
We deduce from the obtained results that our 2. High Rate Ultra Wideband PHY and MAC Standard,
Ecma International Std. 3rd ECMA-368, (2005).
proposed approach based on the appropriate 3. Batra, A., Balakrishnan J., Aiello R., Foerster J. and
choice of MCS rate is a prominent solution for Dabak A.: Design of a multiband OFDM system for
improving the EE through energy per bit realistic UWB Channel Environments. In: IEEE
diminishing since the solution proposed in [19] Transaction on Microwave Theory and Techniques,
vol. 52, pp. 2123--2138, (2004).
determines each MCS's throughput to decide
4. August N., Thirugnanam R., and Sam Ha D.: An
about the optimal one related to channel's Adaptive UWB Modulation Scheme for Optimization
parameters. In other hand, the proposed of Energy, BER, and Data Rate. In: Proc. 2004 IEEE
algorithm in [20] is about identifying the best Workshop In Ultra Wideband Systems, Japan (2004).
MCS scheme in terms of throughput 5. Liu, Q., Zhou, S., Giannakis, G. B: Cross-layer
combining of adaptive modulation and coding with
maximization for each user. However, we
truncated ARQ over wireless links. In: IEEE
remark from the obtained results that this is not Transaction on Wireless Communication, vol. 3, pp.
the best since it doesn't ensure energy per bit 1746--1755. (2004).
diminishing. By this way, we conclude that our 6. Ngoc, P. L., Tran, L. C., Safaei, F., Varma, V. S.:
method is a prominent approach following the Energy efficiency analysis of antenna selection multi-
input multi-output automatic repeat request systems
47
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 42-48
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
48
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 49-59
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
ABSTRACT
49
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 49-59
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
an Analog-to-Digital (A/D) converter with the aim most relevant studies in this domain outlining their
of converting it into a suitable form to be used by main findings.
the microprocessor, (2) the microprocessor Recently, Sabor et al. [5] made a comprehensive
converts the signal into digital data depending on state-of-the-art of the hierarchical-based routing
how it is programmed, and finally (3) it sends the protocols for Mobile Wireless Sensor Networks
information to the network by using a transceiver. (MWSNs). In this study, they classified the
This information is shared between other sensor protocols into two categories: classical-based
nodes and used as input for a distributed routing and optimized-based routing, and analyzed
management system. each group separately based on different metrics
Recent developments in WSNs have focused on such as control manner, network architecture,
several Quality of Service (QoS) parameters such as mobile element, mobility pattern, and applications,
energy, scalability, reliability, accuracy, among others. They also evaluated and compared
adaptability, fault tolerance, and security. Routing the considered protocols using three more
in WSN is considered as one of the most important parameters they are delay, network size, and energy
challenges that highly affects these parameters. efficiency. They argued that, the comprehensive
This is due to the fact that several network study of this set of protocols can be helpful for
constraints might exist; these constrains stem from designers of MWSNs in selecting appropriate
the nature of the application, network architecture, hierarchical routing protocols for a specific
and route establishment. Usually, WSNs aim at application. The same work was done in [6], but
minimizing the overall power consumption by here the authors adopted another taxonomy: flat
employing the appropriate routing protocol to meet and hierarchical based routing protocols. They
these constrains. In this paper, we provide a compared them based on network structure, state of
performance comparative study of four well-known information, energy-efficiency, and mobility.
routing protocols developed for WSNs, and analyze Finally, they gave recommendation about their
the simulation results according to several QoS usages and some insights into the enhancements
metrics. that can be done to improve some of them.
The rest of the paper is structured as follows. First, Al-Karaki et al. [7] classified routing protocols for
we provide a review of related works in Section 2. WSNs into three categories based on the underlying
In Section 3 we give an overview of the four network structure: flat, hierarchical, and location-
protocols that are the subject of this paper. In based routing. They went further and classified
Section 4 we present the simulation testbed and the them into multipath-based, query-based,
common parameters, and in Section 5 we negotiation-based, QoS-based, and coherent-based.
summarize the simulation results and make a They highlighted the advantages and some
comparison based on several QoS metrics. Finally, performance issues of each routing protocol. While,
in Section 6 we provide some conclusions and shed in [8], the routing protocols discussed are classified
the light on some insights about future works. into seven categories they are: Data centric routing,
Hierarchical routing, Location based routing,
2 RELATED WORK Negotiation based routing, Multipath based routing,
Quality of Service (QoS) routing and Mobility
As noted in the introduction section, in recent years based routing, and compared them based on several
many routing protocols have been proposed by parameters such as power consumption, scalability,
researchers worldwide to improve the performance mobility, optimal routing and data aggregation. Jain
of WSNs. Despite the availability of large et al. [9] surveyed the existing cluster-based
proposals, routing in WSN is still a big challenge protocols for both static and mobile WSNs, and
since WSNs are designed for specific applications, made a comparative study among them based on
and every application has its own constraints that several features such as basic assumptions, working
must be considered when designing efficient environment, advantages, limitations, and working
routing protocols. Several comparative studies have style. Location-based routing protocols that support
been carried out by many researchers in order to the sink mobility were revised in [10]. The authors
analyze the performance of the current WSN of this paper classified this category of protocols
routing protocols. In this section, we summarize the into backbone-based and rendezvous-based
approaches according to their network structures,
50
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 49-59
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
and discussed the advantages and disadvantages of 100m), the number of nodes in the grid (sparse grid:
each one of them. 70 in the small grid vs. condensed grid: 1000 nodes
Daesung et al. [11] performed a performance study in the big grid), and the placement of the main BS
for three different types of routing protocols: (in the middle of the grid vs. on the border of the
LEACH, PEGASIS, and VGA using Sensoria grid). The aim of this study is to test the behavior of
Simulator [12]. Power consumption and overall these routing protocols when applied to different
network performance were used as the QoS metrics. scenarios resulted from varying these parameters.
Based on the simulation results, the authors argued
that PEGASIS outperforms the two others, then 3 THE EVALUATED ROUTING
LEACH and finally VGA which has the worst PROTOCOLS
power consumption when the sensing range is
limited, and the best when the sensing range is In this section, we give a brief overview of the four
increased. Bansal et al. [13] analyzed only LEACH WSN protocols analyzed in this work, namely
and PEGASIS based on total energy consumption, LEACH, DEEC, M-GEAR and EESAA. For more
overheads, and sensors lifetime. They argued that information, please refer to the original papers.
PEGASIS outperforms LEACH in terms of network
lifetime, communication overhead and the 3.1 LEACH Protocol
percentage of node deaths. In addition, it offers an
The Low Energy Adaptive Cluster Hierarchy
extended lifetime of the network due to its energy
(LEACH) is the most popular WSNs routing
efficiency. For large networks, the early death of
protocol, which was proposed by Heinzelman et al.
nodes reduces the network stability in LEACH as
[16] many years ago, with the overall aim of
compared to PEGASIS.
reducing power consumption. LEACH works in a
Latif et al. [14] analyzed the performance of four
completely distributed way, and does not need to
routing protocols: LEACH, TEEN, SEP and DEEC
acquire global knowledge of the network. The main
based on a proposed mathematical model they
functions of LEACH can be grouped into two
developed. For validation purposes, they performed
phases: (1) in the setup phase several tasks are
analytical simulations in MATLAB by choosing a
performed related to the organizing of the network
number of performance metrics: number of alive
into clusters, i.e., LEACH is based on the concept
nodes, number of dead nodes, number of packets
of a hierarchical clustering, it divides the entire
delivered, and number of cluster heads. The
network into several clusters and assigns one node
simulation results showed that DEEC outperforms
for each to be the Cluster Head (CH), CH
the others by providing feasible optimal results
advertisement, and creating transmission schedule
against a set of constraints of their suggested
for the entire network, and (2) a steady-state phase
mathematical model. Guangjie et al. [15] conducted
that consists of data aggregation (or fusion),
different set of experiments in MATLAB to analyze compression, and transmission to the BS. LEACH
the performance of five routing protocols: reduces energy consumption by (1) minimizing the
EDFCM, MCR, EEPCA, LEACH, and SEP. The communication cost between sensor nodes and their
simulation results showed that the first three CHs, and (2) shutting down non-CH nodes as much
protocols are better in terms of the time of the first as possible. A CH is responsible for data collection
nodes death and the total number of packets from the entire sensor nodes belonging to the cluster
successfully delivered to the BS. under its control, and then it makes the necessary
The research presented in this paper differs from processing tasks on the data before sending it to the
these works in two main aspects: (1) we have BS. LEACH uses a single-hop routing where each
selected to analyze a different set of protocols, node can directly transmit to the CH and the BS.
namely: LEACH, DEEC, M-GEAR, and EESAA, Therefore, it is not efficient in large networks.
because we felt they have comparable features plus Furthermore, the concept of assigning clustering
they represent a long time span that covers a wide dynamically brings additional overhead such as CH
spectrum of protocols, (2) we implement the four changes and advertisements, which can minimize
routing protocols using MATLAB simulation the gain in energy consumption. While LEACH
platform, but also we create six different scenarios helps the sensor nodes within their clusters to
in a controlled environment by varying three diminish their energy slowly, the CHs consume a
parameters: the grid area (50m x 50m & 100m x larger amount of energy when they are located at
51
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 49-59
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
the farthest point from the BS. Also, LEACH network are grouped into two regions and nodes
clustering terminates in a finite number of within each region are further grouped into clusters.
iterations, but does not guarantee a good CH The first two regions are referred to as non-
distribution and assumes uniform energy clustered regions, while the last two regions are
consumption for CHs, i.e., it adapts a randomized referred to as clustered regions. In the third phase
rotating mechanism of high-energy CH position (CH selection) cluster heads are selected in the two
rather than making a selection in a static manner, to clustered regions, for that purpose, the protocol uses
give an equal chance to all sensor nodes to become the LEACH cluster head selection approach. After
CHs. the selection of CHs, each CH creates a TDMA-
based time-slots for its member nodes, so that each
3.2 DEEC Protocol node can transmit its data in its time-slot and switch
to an idle mode for the rest of the time, this phase is
Distributed Energy-Efficient Clustering (DEEC) called (scheduling phase). The final phase of the
protocol [17], is an adaptive clustering protocol protocol is operation (the steady-state phase), in
used in heterogeneous wireless sensor networks. It which all sensor nodes transmit their sensed data to
assumes that in a WSN there are several types of its cluster heads, which in turn receive and
nodes of different initial energy levels, e.g., a WSN aggregate the data of all its member nodes and send
with three types of nodes of initial energy levels is it to the gateway node. After collecting the data
a three-level heterogeneous network. In DEEC, from all CHs, the gateway node aggregates the data
every sensor node independently elects itself as a and sends it to the BS.
CH, and the probability for a node to be a CH is
calculated based on two parameters: (1) the ratio 3.4 EESAA Protocol
between the residual energy of the node, and (2) the
average energy of the entire network. Therefore, a Energy Efficient Sleep Awake Aware intelligent
node with high initial energy and residual energy sensor network routing protocol (EESAA) [19] uses
acquires more chances to become a CH. To control the concept of pairing and equipping the nodes with
the energy consumption of sensor nodes by means Global Positioning System (GPS) capabilities. In
of adaptive approach, DEEC uses the average addition to the enhanced cluster head selection
energy of the network as the reference energy. technique, the protocol was designed to improve
Thus, DEEC does not require any global knowledge network stability period and prolong the network
of energy at every election round. Unlike LEACH, lifetime.
DEEC can perform better in multi-level Sensor nodes use their GPS to calculate their own
heterogeneous WSNs. locations and send them to the BS, where in the first
round; cluster head selection takes place using
3.3 M-GEAR Protocol LEACH CH selection technique. When BS receives
the nodes location information, it groups the nodes
Gateway-Based Energy-Aware Multi-Hop routing into pairs (create couples) based on the minimum
protocol for WSNs (M-GEAR) [18] uses a BS and distance from each other in their intra-cluster
a gateway node. It divides the nodes into four transmission range, and also they should belong to
regions. The operation of the protocol consists of the same application type. Then it broadcasts this
five phases, in its initial phase the BS broadcast a pairing information to all nodes in the network for
HELLO packet, and all nodes in the network replay each node to become aware its coupled node.
indicating their locations. Then, the BS can build During a single communication interval, paired
the nodes data table by saving each nodes id, nodes switch between awake and sleep modes. In
location, distance from the BS and gateway node any given pair, at first the node closer to the BS
and residual node energy. In the second phase switch into awake mode (active mode) and is
(setup), the protocol divides the network into four responsible to collect data from its surroundings
regions, the first one for nodes near the BS (nodes and send it to the CH, while its coupled node
in this region can send data directly to the BS), the remains in sleep mode during this period. In the
second region for nodes near the gateway node next round, nodes in active mode switch into sleep
(nodes in this region can send their data to the mode while its couple becomes awake, and so forth.
gateway node which in turn aggregates the data and
forwards it to the BS). The rest of the nodes in the
52
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 49-59
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
Unpaired nodes remain in active mode in all the cluster heads per round, and finally (iv) the number
rounds until they die. of packets that went through the cluster heads. Next,
The selection of CHs after the first round depends we make a comparative study of their performance
on remaining energy of active nodes only, so in the in six scenarios in order to study the impact of
start of each round nodes transmit information changing the position of the sink node, network
about their energy to CH, which in turn calculates size, and nodes distributions.
the energy of every node in its cluster and distance
from each node and select CH for the next round. In Table 1. Network simulation parameters.
this regard, the selected CH broadcast an Parameter Value Description
advertisement message that can be heard by every Eelec 50 nj/bits power needed to run the
awake node to select their CH according to the transmitter or receiver
received signal strength indication of those circuitry
advertisements. Finally, nodes in active mode are
the only nodes allowed to transmit their sensed data Efs 10 pj/bit/m2 % Transmit Amplifier
type
during their TDMA-slots while nodes in sleep mode
turn their transceivers off to save energy. EDA 50 nj/bit/packet Data Aggregation
Energy
4 SIMULATION TESTBED
E0 0.5 J Initial Energy
As mentioned earlier, we have used MATLAB to Number of 5000 --
run our simulations. MATLAB is an ideal choice rounds
Emp 0.013pj/bit/m4 % Another Transmit
for WSN simulations since it gives accurate results Amplifier type
and includes built-in capabilities. For each
simulation scenario, we designed a network with a
5 SIMULATION RESULTS AND DISCUSSION
predefined area and then distributed a given number
of nodes randomly within that area. Each one of the As previously mentioned, in the first two scenarios
four protocols has been simulated with 6 different we built a small-area scattered network of 70 nodes
scenarios (as shown in Table 2), but in all of these dispersed randomly in 50m * 50m simulation area
scenarios we used common fixed radio parameters. with a BS located at the center and the border
Also, each node was considered dead when its respectively. In the third and fourth scenarios, we
internal energy approaches zero. To simulate built a dense network of 500 nodes distributed
message exchanges between nodes we used a fixed randomly in 100m * 100m area with BS located at
message size of 4000 bit, and finally, it is worth the center and the border of the area respectively,
mentioning that we assumed the network to be free and finally in the fifth and sixth scenarios, we built
of collisions. Table 1 lists the network parameters a highly- dense network of 1000 nodes distributed
used in the testbed scenarios. randomly in 100m * 100m area with BS located at
The four WSN routing protocols surveyed in this the center and the border of the area respectively.
study are already built-in MATLAB, which we Table 2 summarizes the characteristics of these six
believe is a very good thing since it enables us to simulation scenarios.
get reliable and consistent results. We designed the
testbed to simulate six scenarios for of each one of Table 2. The six simulated scenarios.
the four protocol to study the impact of changing
Num. of BS
the sink node position, the impact of having Scenario Area Property
nodes location
different network sizes and densities, and the S1 70 50m x 50m Center Sparse
number of nodes and their distributions. In the S2 70 50m x 50m Border Sparse
following subsections, a summary about the setup S3 500 100m x 100m Center Dense
variables on each scenario will be listed, then we S4 500 100m x 100m Border Dense
provide the main differences between the four
S5 1000 100m x 100m Center Hi-Dense
protocols LEACH, DEEC, M-GEAR and EESAA
S6 1000 100m x 100m Border Hi-Dense
using four parameters: (i) the number of the dead
nodes per round, (ii) the number of packets
successfully received by the BS, (iii) the number of
53
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 49-59
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
5.1 Number of Dead Nodes per Round protocols, mainly M-GEAR, DEEC, and EESAA.
In M-GEAR, the number of dead nodes is almost
The first metric analyzed in this study is the the same when applied in low dense networks (S1
number of dead nodes. It is considered as one of the and S2), but it behaves badly in the case of bigger
mostly relevant metrics used by researchers to networks (S3 to S6), therefore when using M-
evaluate the performance of a given routing GEAR, it is better to locate the BS at the center of
protocol, because it affects other important metrics, the network to get the maximum performance in
specifically network lifetime and network stability terms of network stability and lifetime. The same
period. Network lifetime is defined as the time finding is almost valid for the EESAA protocol.
duration between the network process initialization Here, EESAA gives best performance measure
and the expiry of the last alive node in the network, when the BS is located at the center of network in
while the stability period can be defined as the time large networks, this is contrary to the M-GEAR
duration between the starting of the network which is slightly better small networks when
process and the expiry of the very first node in the locating the BS at the network borders. Finally, in
network (i.e. the time when the first node dies). DEEC locating the BS at the center of the network
Figure 2 shows the simulation results of the number area gives higher instability period compared to the
of dead nodes vs. the number of rounds of the six BS located at the borders regardless of the network
scenarios. It is worth mentioning here that we put size. It is worth mentioning that the performance
the simulation results of the six scenarios in one big results of the four protocols in this plot seem to be
plot to facilitate the comparative process as we consistent between the large grid-dense cases (S3 &
forward from Scenario 1 to Scenario 6. From the S4) and the large grid-very dense cases (S5 & S6),
figure, we derive the following findings: (1) for all which is basically an indication that even if we
scenarios, it is clearly observable that LEACH gives double the number of nodes in the large grid (100m
the minimum stability period then comes M-GEAR, x 100m), we still get very similar results. Table 3
DEEC, and finally EESAA which has the best shows the rounds number of the first dead node.
stability measure among the others; (2) Also, Here, the numbers emphasize our findings obtained
EESSA protocol has the maximum network from Figure 2. The numbers show that ASSEE
lifetime, and gives incredible results compared to outperforms all other routing protocols with about
others, which almost have the same value with 100% stability period and network lifetime in all
some variants, especially for DEEC; (3) We scenarios; here we can clearly see that all numbers
observe that changing the location of the BS has in the ASSEE column are larger than 2170 with
significant effects on the performance of some case S6 being the worst with 2171 which is way
54
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 49-59
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
better than the other three. Then comes DEEC, protocol is doing its job properly or not. Figure 3
LEACH, and M-GEAR, and in that order; DEEC in shows the simulation results of the six scenarios.
the range 930 (worst performance in S4) and 1390 Several findings are extracted as follows: (1) in the
(best performance in S1), LEACH in the range 892 first two scenarios (S1, S2), DEEC outperforms the
(worst performance in S6) and 984 (best others in the number of packets successfully
performance in S3 & S4), and finally M-GEAR in delivered to the BS, then comes M-GEAR, EESAA,
the range 521 (worst performance in S6) and 1070 and finally the LEACH which has the minimum
(best performance in S1). This is evidence that number of packets received by the BS. Whereas, in
DEEC and M-GEAR are badly affected by the scenarios S3 and S4, EESAA outperforms all of
number of nodes in the grid. Furthermore, M- them, while M-GEAR gives the minimum number.
GEAR has almost the same performance measure Both DEEC and LEACH are located at the middle;
as DEEC for the first two scenarios, while it gives (2) regarding the effect of the location of the BS, we
the minimum performance results in the large and find that in LEACH the number of packets
dense grids. delivered to the BS is almost the same whether the
BS is located at the center or at the borders of the
Table 3. Rounds number of the first dead node. network, with a slightly higher number in the
Scenario LEACH DEEC M-GEAR EESAA former case. DEEC gives the best results of packet
S1 912 1391 1070 2236
delivery ratio, mainly when the BS is located at the
center. Like DEEC, M-GEAR does the best job
S2 978 1113 1053 2283
when the BS is located at the center. Finally, there
S3 984 994 630 2233 is a significant effect of the location of the BS on
S4 983 934 526 2182 the functioning of EESSA. Here, we observe that
S5 966 1042 597 2336 EESAA gives better results when the BS is located
S6 892 1018 521 2171 at the center when deployed in large networks (S3
and S4), whereas in the case of small networks (S1
5.2 Packets to the Base Station and S2) its highly recommended to put the BS at the
borders to get the maximum benefits of this
In this section, we study another performance protocol.
metric which is the number of data packets
successfully received by the BS. This metric gives 5.3 Cluster Heads Formation
good measures about the reliability of the network,
i.e., it presents a good indicator whether the routing
55
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 49-59
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
The third performance metric considered in this which are located near the BS. Nodes in this region
study is the stability of the cluster heads formation can send data directly to the BS; The second region
process. This process plays an important role in the encompasses a set of nodes that are close to the
stability of the whole network. gateway node. Nodes in this region send their data
Unlike the aforementioned performance metrics, a to the gateway node, which in turn aggregates the
protocol having a large number of cluster heads per data and forward it to the BS; the remaining nodes
round does not mean that the protocol behaves are grouped into two regions, and nodes belonging
better than of those having small ones. Here, we to a group are further split into more clusters. The
perform the analytical study considering that the first two regions are referred to as non-clustered
cluster heads per round for a particular protocol is regions, while the last two are referred to as
much closer to the optimal situation when this clustered regions. This process explains the
number follows a uniform behavior of M-GEAR protocol in creating fewer
distribution within the network lifetime. Based on cluster heads in the large grids compared to the
this definition, Figure 4 shows the simulation small ones.
results of the four protocols. Here, we extract the
following findings: (1) in all scenarios, its obvious 5.4 Number of Packets at Cluster Heads
that EESAA has the highest stability in cluster head
formation process as the number of cluster heads As the fourth and last performance metric in our
created is uniformly distributed within the whole study, we will study the impact of the total number
network lifetime with an average of 4 cluster of packets that went through the cluster heads on the
heads/round. DEEC has low cluster formation network. We can consider this parameter as an
stability in all scenarios; in small grid networks approximation of the centralized nature of the
LEACH shows low cluster formation stability, but protocol and the overhead traffic imposed by the
in large grid networks it shows moderate stability protocol on the cluster-heads. Figure 5 presents the
level. M-GEAR also has a moderate level of cluster simulation results of the four protocols. Here, we
stability for all scenarios. (2) contrary to what is can extract the following findings: (1) in the
expected, M-GEAR creates a small number of small/sparse grid (S1 & S2), all protocols are
cluster heads per round as the network becomes generating big and noisy traffic at the cluster heads,
larger (S3 to S6). This is due to some design issues while they perform better when the grid is big and
behind this protocol. In the setup phase of the dense where we can see smoother and smaller plots,
operation of this protocol, it divides the network (2) it seems that the BS location being at the center
into four regions; the first one consists of nodes or at the edge does not really affect this parameter.
56
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 49-59
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
In this plot, we can see that (S3 & S4) are almost while DEEC performing better in columns 1 and 2
identical and the same can be observed for (S5 & and high-order grids. LEACH took the last place
S6), (3) we can see that the M-GEAR is giving best with Weak to Moderate performance in all
performance here with minimum overhead, scenarios.
followed by EESAA and last come LEACH and
DEEC. We can attribute this good performance of 6 CONCLUSION AND FUTURE WORK
M-GEAR to its smart multi-hop hierarchy such that
in level zero nodes can send to the BS directly, and In this paper, we surveyed and compared the
in level 1 nodes can send to the gateway then to BS, performance of four well-known routing protocols
and only in the outer regions cluster heads are which were mainly developed for WSNs, namely
involved. The EESAA gives good performance LEACH, DEEC, M-GEAR, and EESAA. The
here due to its smart pairing mechanism which also performance of the four protocols was tested and
reduces this kind of overhead, (4) the last evaluated in four performance metrics and under six
observation on this plot is the shape of the curve different scenarios. The empirical results have
(especially in the big grid cases); it starts semi- shown that the performance of the protocols differs
linear and then a sharp drop happens in the middle as we go from small grid and sparse networks to
when the nodes start to lose their energy and die. larger grid dense networks. Some protocols gave
Here we can see that EESAA gave the longest best performance when applied at small grid
survival time among the other which all dropped to networks, and moderate or even low performance
zero between rounds 2500 to 3000. when applied at large grid networks and vice versa.
In summary, Table 4 presents all these findings in a Furthermore, we have found that the location of the
more readable way. In this table, we use keywords BS can have a significant effect on networks
to evaluate these protocols targeted at the three stability. Some protocols provide high stability and
previously discussed metrics. These keywords are long network lifetime when the BS is located at the
High, Moderate and Low. High means highest center of the network, and others perform better
performance, while Low refers to the lowest when the BS is located at the borders of the
performance. If we take the average of all six network. As illustrated in Table 3, EESAA has the
scenarios and all four parameters (total 24 entries in highest performance measures in all scenarios.
the table for each protocol), it is obvious that DEEC and M-GEAR come next with a moderate to
EESAA is the winner with best performance in all high levels, with one outperforms the other in some
categories, next come M-GEAR and DEEC, with cases, and finally comes the LEACH which has the
M-GEAR performing better in columns 3 and 4, lowest performance measures.
57
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 49-59
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
Based on these findings, we will continue to and Future Directions, Wireless Communications and
investigate the performance of these routing Mobile Computing, vol. 2017, Article ID 2818542, 23
pages, 2017. doi:10.1155/2017/2818542.
protocols in more complicated scenarios and will
6. G. S. Sara and D. Sridharan, Routing in mobile wireless
try to combine the best features in them (which led sensor network: a survey, Telecommunication Systems,
to their good performance) specially EESAA and vol. 57, no. 1, pp. 5179, 2014.
DEEC, to design a new (hybrid) routing protocol 7. J. N. Al-Karaki and A. E. Kamal, Routing techniques in
and test it as well. wireless sensor networks: a survey, in IEEE Wireless
Communications, vol. 11, no. 6, pp. 6-28, Dec. 2004. doi:
REFERENCES 10.1109/MWC.2004.1368893.
H. S. Bazzi, A. M. Haidar, and A. Bilal, Classification
1. I. F. Akyildiz, Weilian Su, Y. Sankarasubramaniam and of routing protocols in wireless sensor
E. Cayirci, A survey on sensor networks, in IEEE network, Computer Vision and Image Analysis
Communications Magazine, vol. 40, no. 8, pp. 102-114, Applications (ICCVIA), 2015 International Conf. on,
Aug 2002. doi: 10.1109/MCOM.2002.1024422. IEEE, 2015.
2. J. Yick, B. Mukherjee, and D. Ghosal, Wireless sensor 8. M. Patil and R. C. Biradar, A survey on routing
network survey, Computer Networks, vol. 52, issue 12, protocols in wireless sensor networks, 2012 18th IEEE
pp. 2292-2330, August 2008, ISSN 1389-1286, International Conf. on Networks (ICON), Singapore,
http://dx.doi.org/10.1016/j.comnet.2008.04.002. 2012, pp. 86-91.
3. D. Culler, D. Estrin, and M. Srivastava, Overview of doi: 10.1109/ICON.2012.6506539.
sensor networks, Computer Magazine,IEEE, vol. 37, no. 9. S. R. Jain and N. V. Thakur, Overview of cluster based
8, pp. 41-49, August 2004. routing protocols in static and mobile wireless sensor
4. C. Y. Lin, S. Zeadally, T. S. Chen, and C. Y. Chang, networks, in Information Systems Design and Intelligent
Enabling cyber physical systems with wireless sensor Applications: Proceedings of Second International
networking technologies, International Journal of Conference INDIA 2015, Volume 1, pp. 619626,
Distributed Sensor Networks, vol. 2012, Article ID Springer India, New Delhi, India, 2015.
489794, 21 pages, 2012. doi:10.1155/2012/489794 10. K. Daesung, K. Dongkyun, P. Hyeon, and Y. Seung-mok,
5. Nabil Sabor, Shigenobu Sasaki, Mohammed Abo- Performance evaluation of routing protocols for wireless
Zahhad, and Sabah M. Ahmed, A Comprehensive sensor networks in military scenarios, 2011 Third
Survey on Hierarchical-Based Routing Protocols for International Conf. on Ubiquitous and Future Networks
Mobile Wireless Sensor Networks: Review, Taxonomy, (ICUFN), Dalian, 2011, pp. 101-106.
doi: 10.1109/ICUFN.2011.5949143.J. Wang, X. Yang, Z.
58
International Journal of Digital Information and Wireless Communications (IJDIWC) 7(1): 49-59
The Society of Digital Information and Wireless Communications, 2017 ISSN: 2225-658X (Online); ISSN 2412-6551 (Print)
59
International Journal of
DIGITAL INFORMATION AND WIRELESS COMMUNICATIONS
The International Journal of Digital Information and Wireless Communications aims to provide a forum
for scientists, engineers, and practitioners to present their latest research results, ideas, developments
and applications in the field of computer architectures, information technology, and mobile
technologies. The IJDIWC publishes issues four (4) times a year and accepts three types of papers as
follows:
1. Research papers: that are presenting and discussing the latest, and the most profound
research results in the scope of IJDIWC. Papers should describe new contributions in the
scope of IJDIWC and support claims of novelty with citations to the relevant literature.
2. Technical papers: that are establishing meaningful forum between practitioners and
researchers with useful solutions in various fields of digital security and forensics. It includes
all kinds of practical applications, which covers principles, projects, missions, techniques,
tools, methods, processes etc.
3. Review papers: that are critically analyzing past and current research trends in the field.
Manuscripts submitted to IJDIWC should not be previously published or be under review by any other
publication. Original unpublished manuscripts are solicited in the following areas including but not
limited to: