Professional Documents
Culture Documents
International Journal of
ISSN 2225-658X (Online)
DIGITAL INFORMATION AND WIRELESS COMMUNICATIONS
Volume 3, Issue 2
2013
TABLE OF CONTENTS
Original Articles
PAPER TITLE
AUTHORS
Brendan Flanagan, Takahiko
PEER KNOWLEDGE ASSISTED SEARCH USING COMMUNITY SEARCH Suzuki, Jun Zeng, Chengjiu Yin,
LOGS
Toshihiko
Sakai,
Kiyota
Hashimoto, Sachio Hirokawa
PAGES
141
149
A MATLAB BASED
LABORATORY
169
CELLULAR
MOBILE
Driss
181
Lakshmi
Ramachandran,
AN INTERSECTION BASED TRAFFIC AWARE ROUTING WITH LOW
Sangheethaa Sukumaran, Surya
OVERHEAD IN VANET
Rani Sunny
190
197
M. Demba
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 141-148
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
Brendan Flanaganb, Takahiko Suzukia*, Jun Zengb, Chengjiu Yina, Toshihiko Sakaib,
Kiyota Hashimotoc and Sachio Hirokawaa
a
Research Institute for Information Technology, Kyushu University, Japan
b
Graduate School of Information Science, Kyushu University, Japan
c
School of Humanities and Social Sciences, Osaka Prefecture University, Japan
*6-10-1, Hakozaki, Higashi, Fukuoka, Japan
*suzuki@cc.kyushu-u.ac.jp
ABSTRACT
Search engines have become an increasingly
important educational tool. By just typing a
few keywords into a search engine text-box,
finding information has become an easy
exercise. However, some people cannot use
search engines effectively for various
reasons. Students may not know how to
choose proper keywords. They may have
difficulty in selecting relevant information
from millions of search results. In this paper,
we propose a search engine system that
shares search queries and browsing history
among students. We refer to the search
queries and browsing history as community
search logs. This system analyzes these
community search logs, and shares the
students knowledge and experience with
each other. Our purpose is to help students
to improve their searching efficiency,
especially those who are not good at
searching.
KEYWORDS
Community Search Logs, Search Engine,
Knowledge and Experience Sharing, Peer
Knowledge Assisted Search.
1 INTRODUCTION
Along with the development of the
Internet and search engine technology,
search engines have become an
increasingly important educational tool.
Researchers have indicated that allowing
students to learn as active and self-
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 141-148
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
experiments, if School Science Fair
Projects is entered instead of science
projects, this will return more specific
information.
Users may also have difficulty in
selecting the relevant information from a
large number of search results. Although
most search engines employ methods to
rank results in order to provide the
best results first, not all search results
that are ranked high are relevant to the
users needs.
We propose a search engine system that
shares search queries and browsing
history among a group of users. We refer
to the search queries and browsing
history as community search logs.
This system analyzes these community
search logs, and shares the users
knowledge and experience with each
other. Our purpose is to help students to
improve their searching efficiency,
especially those who are not good at
searching.
This paper is structured as follows:
Section 2 is a brief introduction of the
related works. In section 3, we explain
the preparation of data and a preliminary
experiment. Section 4 describes the
proposed system in detail. Section 5
describes our conclusions and future
work.
2 RELATED WORKS
Previous research has been conducted on
using the access logs of Web pages.
Toda et al. [6] focused on Web pages
that were often visited by many users in
order to improve the efficiency of Web
browsing routines as well as the
discovery of new information. In order
to evaluate the importance of the pages
for recommendation, they analyzed the
time and frequency of users who visited
particular pages in access logs. This was
142
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 141-148
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
SNS search. By seamless use of various
activities performed while searching and
mutual comments, users can share
information on problems or new
knowledge.
3 DATA PREPARATION AND
PRELIMINARY EXPERIMENT
Two kinds of data are utilized in the
proposed system: a database containing
search targets that we refer to as search
data, and the users search query and
browsing history that we refer to as
community search logs. In this
section, we will introduce the procedure
used to collect both of these kinds of
data.
3.1 Collecting Search Data
Information on papers about mobile
learning was collected as the search
data. The keywords mobile learning
was used to conduct a search using
Scopus 1 , which is the worlds largest
abstract and citation database of peerreviewed literature. As a result, we
collected the information of 13353
papers. Each paper contains the
following components:
Title of the paper
Authors of the paper
Departments, organizations and
countries of the authors
Publisher
Publication date
Keywords
Abstract
A frequency file of the words appearing
in those components was generated.
Finally, this frequency file was then used
to build an index using GETA2, which
1
2
http:// www-scopus-com
http://geta.ex.nii.ac.jp/
143
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 141-148
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
some novice users might complain that
they found it difficult to utilize those
various search hints.
3.3 Collecting Community Search
Logs
The search queries and browsing history
of the experimental group were collected
as the log data. As a result, 84 search
queries and 192 browsed web page
histories were collected. Figure 1 shows
the segments of the search queries. Each
search query contains the following:
The number of users who used
the query
The keywords of the query
Figure 2 shows the segments of log data.
Each entry of the log contains the
following:
The IP address of the PC used by
the participants
The date and time when the
participants browsed the search
results
Paper number (1~13353)
Keywords used by participants
22 mobile learning
17 mobile learning y:2010
9 mobile learning device y:2010
7 mobile learning y:2001
7 mobile learning c:Taiwan
144
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 141-148
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
However, at the initial stage of the
research only the log data mentioned in
section 3 was utilized, and no new log
data was created while users were using
the proposed system.
4.2 The Functions and User Interface
Of The Proposed System
B
C
145
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 141-148
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
Publisher
Publication date
Keywords
Abstract
There is a like button at the top of
frame C as shown in Figure 6. Users can
express their interest in the search result
document by clicking the button.
146
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 141-148
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
3.
2.
147
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 141-148
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
13. Y. Chuang & L. Wu, User-Based
Evaluations of Search Engines: Hygiene
Factors and Motivation Factors, Proc. of
40th
Annual
Hawaii
International
Conference on System Sciences, 2007, pp.110.
148
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 149-168
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
KEYWORDS
Energy efficiency, femtocell, interference
coordination, LTE-Advanced, resource
reuse, scheduling, spectral efficiency.
1 INTRODUCTION
This research focuses on LTEAdvanced systems and femtocell
technologies, the cutting-edge systems
and technologies of the next generation
mobile communication networks. The
LTE-Advanced system is under the
design and investigation phase, with
several critical issues yet to be addressed
and left as open research problems. One
of the critical issues is to achieve high
data rate to provide heterogeneous
services. However, high data rate
requires highly spectral efficient
networks. The conventional frequency
reuse factor of one in the macrocell is
not sufficient to meet high spectral
efficiency requirements of LTEAdvanced systems. This drives the use
of smaller cells within the coverage of a
macrocell.
Further, high data rate demand of
users for multimedia communication
149
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 149-168
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
causes increase in energy consumption
of radio access networks [1]. Almost
80% of the total power consumptions in
mobile networks is due to radio access
networks [2]. This results in paying out
considerable attention to energy
consumption of the next generation radio
access networks [3]. Design of energy
efficient radio has now become a new
research direction in both industry and
academia [2]. Several approaches have
been proposed to reduce the base station
(BS) energy [6]. Reducing distance
between transmitter and receiver allows
reduction in transmit power to provide
good received signal quality [2] and can
be implemented by deploying femtocells
as second tier in the existing macrocells.
A femtocell is a low power BS,
typically consumes few watts much less
than that of a macro BS, which covers
small area with good signal quality,
typically deployed as the overlaid
network in the coverage of macrocells
[7]. Hence, by serving sufficient amount
of traffic using femtocells, significant
amount of energy consumption can be
reduced. In addition, because of close in
distance of a femto BS to a mobile
station (MS), the received signal strength
is also improved, which results in
spectral efficiency improvement as well.
Furthermore, scheduling of resources
to a user has significant impact on the
achievable data rate that is normally
performed based on the measurement of
the channel condition between the user
equipment (UE) and the macro BS. The
UE provides the channel information to
the macro BS through uplink
transmissions.
Several
researches
already
addressed
the
resource
scheduling in the macro-femto cell
network architecture based on the
resource reuse among femtocells to
increase the data rates [23]. However,
WORK
AND
150
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 149-168
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
a decentralized resource allocation
scheme where each femtocell is
restricted to a randomly chosen subset of
resources; and Ref. [16] investigated the
degree of access control of femtocells
and its impact on a mixed macro-femto
cell system based on the co-channel
assignment strategy.
Further, Ref. [2] proposed a new
resource allocation strategy to improve
downlink energy consumption. Ref. [3]
showed that introducing sleep mode in
small cells can improve energy
consumption gain. Ref. [7] emphasized
the
importance
of
interference
management in achieving energy
efficiency.
Furthermore, a number of papers have
identified that resource allocation
techniques have the potentiality of
improving
energy
efficiency
significantly [1, 4, 5, and 6].
However, very few or none researches
have demonstrated explicitly how the
improvement in both spectral efficiency
and energy efficiency can be achieved
with resource scheduling on the same
platform using femtocells. In this paper,
we propose a novel radio resource reuse
and interference coordination technique
employed to the femto tier in order to
improve both energy and spectral
efficiencies.
In doing so, we develop a novel
resource scheduling algorithm on top of
the PF scheduling strategy to minimize
the implementation cost and complexity.
Furthermore, we investigate the effect of
interference on system performances and
develop an interference coordination
strategy to make the scheduler more
realistic. We also investigate the fairness
performance for macro users and femto
users separately and develop a new
mechanism to improve the fairness
performance of the system as a whole.
151
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 149-168
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
Path
loss:
UE to
macro
BS
Path
loss:
UE to
femto
BS
UE is
outside
PL (dB) =15.3 +
37.6log10R, R in m
UE is
inside an
apartment
PL (dB) =15.3 +
37.6log10R + Low, R in m
Dualstripe
model
PL(dB) =
127+30log10(R/1000), R
in m.
3 1010
3.5 10 9
=> 2 = 4.28cm.
4.28 cm
833 cm s
152
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 149-168
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
1
1
~
Rk ,i ( )(1 ) Rk , j ,i ( ), k k i, j ( )
~
tc
tc
Rk ,i ( 1)
1
~
Rk ,i ( )(1 ),
k k i, j ( )
t
c
UE
(2)
th
10
10
10
A Macro
Cell Site
10
10
10
10
A Dual Stripe
Apartment Block
10
m
th
in the i CC at the j th RB in time ;
~
Rk ,i ( ) is the average throughput in the
th
past of the user k in the i CC in time .
The throughput of the user ki , j ( ) is
Rk , j ,i ( )
k i, j ( ) arg max ~
R
(
)
k
k ,i
(1)
user in the
Spectral Efficiency
Peak spectrum
Uplink
15
Downlink
30
153
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 149-168
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
efficiency (bps/Hz)
Average spectrum
efficiency
(bps/Hz/cell)
Cell-edge spectrum
efficiency
(bps/Hz/cell/user)
1.2 ~ 2.0
2.4 ~ 3.7
0.04 ~ 0.07
0.07 ~
0.12
154
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 149-168
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
An RB is considered to be reused only
when that RB is assigned to a femto
user by the PF scheduler in a TTI.
2) Intra-tier Interference Coordination
Strategy: For macro tier, since RB
allocations among macro users are
always orthogonal to each other, there is
no need for the interference coordination
among macro users. However, for femto
tier, both Inter-floor modelling and Intrafloor modelling is considered.
a) Inter-floor Modelling: RBs are
equally divided into two groups after
reserving a group of RBs for the fairness
improvement of macro user. Each RB
group is assigned to femto users of the
alternate floors (figure 6).
b) Intra-floor Modelling: An RB is
considered to be reused by another femto
user only when that femto users serving
femto BS is non-adjacent to the already
assigned femto users serving femto BS,
irrespective of the number of RB reuses
considered. The resource reuse distance
between femto BSs must be at least 10
meters, irrespective of femto BS
locations in the same floor, either in the
same stripe or in different stripes.
3.8 Resource Scheduling Algorithm
The resource scheduling algorithm
illustrates a systematic and stepwise
method to represent the proposed
resource reuse and scheduling strategy
and incorporates a number of steps as
follows:
Step 1: reserve a certain percentage of
total RBs of the system bandwidth for
macro users. Step 2: estimate and sort
PM of all users at all RBs and do
mapping of sorted PM to corresponding
users.
Step 3: find RBs that are
scheduled to macro users, UMk where k
is any positive integer denotes a user
index. Step 4: find RBs that are
155
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 149-168
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
non adjacent to all femto users scheduled
in their respective RBs, irrespective of
whether or not they are in the same floor
in the femtocell apartment block. Hence,
in the proposed scheduler, RB 4 and RB
5 can be reused simultaneously to three
additional femto users on top of the
femto users scheduled by the PF
scheduler.
PM
RB 1
RB 2
RB 3
RB 4
. . . . . . . . . . . .N
. . . . . . . . .
RB 5
Nmax
1st
max
PM
UFk
UFk
2nd
max
PM
UFk
UFk
UFk
UFk
3rd
max
PM
UFk
UFk UMk
Mmax
RB
UFk
. . . . . . . . . . .
RB 1
RB 2
RB 3
RB 4
RB 5
UFk
. . . . . . . . . . . .N
. . . . . . . . .
UMk UFk
RB
Nmax
UFk
UFk
UFk UMk
UFk
2nd
max
PM
UFk
UFk
UFk
UFk
UFk
UFk
UFk
UFk
3rd
max
PM
UFk
UFk
UFk
UFk
UFk
UFk
Mmax
UFk
UFk
UFk
UFk
UFk
. . . . . . . . . . .
PF
Proposed scheduler
scheduler
RB 1
RB 2
RB 3
RB 4
RB 5
. . . . . . . . . . . . N. . . . . . . . .
Nmax
UFk
UFk
2nd
max
PM
UFk
UFk
UFk
UFk
UFk
3rd
max
PM
UFk
UFk
UFk
Mmax
UFk
UFk
UFk
RB
UFk
. . . . . . . . .
An RB reused in a TTI
156
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 149-168
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
P CC ( i )
PTRB ( j ) = T ALL
X
(3)
XR
X ALL
100
(4)
3) Spectral Utilization:
The total
number of RBs allocated by the PF
scheduler on a CC in any TTI is given
by X TPF = X All , whereas the total number
of RBs allocated by the MPF scheduler
is given by X TMPF = X All + X TR where X TR
is the number of reuse RBs allocated by
the MPF scheduler in a TTI. Hence,
compared to the PF scheduler, the
spectral utilization improvement factor
of the MPF scheduler is given by,
MPF
U PF
=
X TMPF
X TPF
(5)
PTRB ( j ).H ( j )
Z
(6)
N ( j ) + I z ( j )
z =1
1.38,
k 13
(7)
( dB)
4.4,
22dB
where = 0.6 , is the implementation
157
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 149-168
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
estimate the average throughput per user
during the course of the simulation
runtime. For the PF scheduler, the
PF
average MI ( ) over all RBs over all
users in a CC is obtained, and then the
PF
effective SINR, eff
per RB per TTI is
estimated from the average MI by
inverse mapping as shown in figure 5.
The inverse mapping function between 10 dB and 22 dB SINR can be
mathematically expressed for the PF
scheduler as
PF
eff
( PF
PF
(2
ee =
(9)
Hence, the system level spectrum
efficiency for the PF scheduler per RB
per TTI can be expressed as
) = 10 log 10
-1)
PF
PF
Rsys
= log2 ( 1 + eff
)
(10)
Similarly for the MPF scheduler,
system level spectrum efficiency per RB
per TTI can be expressed as
MPF
eff
( MPF ) = 10 log10
MPF
(2
(11)
where the average MI and the
effective SINR per RB per TTI for the
- 1)
MPF
MPF
[ J/b]
(14)
K
uk
FJ k 1K
2
K uk
(15)
k 1
(12)
158
m
en
th
ei
gh
t:
m
ax
i
m
um
10
flo
or
s
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 149-168
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
A
pa
rt
UE
Floor 3
Floor 2
Floor1
3m
10 m
10 m
31
32
33
34
35
36
37
38
39
40
21
22
23
24
25
26
27
28
29
30
11
12
18
19
20
10
Macro UE
13
An
UE
3
Serving
femto BS
14
15
16
17
Femto UE
10 m
10 m
A Macro
Cell Site
RB
Comparator
CQI
Manager
LA
RB Manager
RB
Directory
Femto RB
Distributor
TD
Scheduler
Instantaneous Throughput
User
Update
RB
Aggregator
2
User Index
Directory
PF Scheduler
PM Generator
Reserved Macro RB
Index Controller
4
User Average
Throughput
Directory
5
Interference Coordinator
Interference
Coordination
Policy Directory
8
7
PM-User Index
Mapping Generator
3/8
User Index
Comparator
OPTION 1:
9
Decision
Support System
RB Reuse
Decision Maker
PF Scheduler
OPTION 2:
MPF Scheduler
RB Allocation
10
159
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 149-168
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
3.11
Simulation
Parameters,
Assumptions and Scenarios
Simulation parameters and assumptions
are adopted from [28] and are listed in
table 3. Table 4 lists the major scenarios
that are considered for the evaluation of
system performances.
4 SIMULATION RESULTS
4.1 Performance of MPF Scheduler
with Interference Coordination
The MPF scheduler can gain advantage
from the multi-user diversity since an RB
can be simultaneously reassigned to a
number of femto users. Figures 8 and 9
respectively present the performance of
the MPF scheduler over the PF scheduler
in terms of average throughput and
fairness. It can be found that the MPF
scheduler outperforms in throughput
performance
whereas
shows
approximately the same in fairness
performance as the PF scheduler.
Note that, in the absence of
interference coordination (figure 8), the
MPF scheduler can obtain higher average
throughput than with considering
interference coordination. However, it
can be obtained only at lower fairness
performance (figure 9). This lower
fairness in the absence of interference
coordination can be explained by figure
10 that shows RB distributions among
users. The distribution of RBs with
interference coordination is more
clustered than in the absence of
interference coordination.
The greater the degree of evenness in
RB allocations among users is achieved,
the higher the value of average fairness
index. Unlike in the absence of
interference coordination, the MPF
scheduler with interference coordination
160
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 149-168
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
from multi-user diversity of macro users.
In addition, since the number of macro
users is lower than that of femto users,
this deviation in RB allocations between
the macro user group and the femto user
group becomes high when the system is
fully loaded.
161
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 149-168
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
allocated to a macro user and the average
number of RBs allocated to a femto user
is reduced. In addition, the mechanism
also improves fairness among macro
users only (figure 15) with little or no
degradation in fairness among femto
users only (figure 16).
162
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 149-168
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
upon the aggregated throughput in a cell.
The aggregated throughput in a cell is
improved as long as the number of femto
users scheduled in a TTI is increased
because of good channel quality, and is
not affected negatively by the interference
coordination in the femto tier or the
frequency of resource reuse to a particular
femto user in a TTI.
However, as given in table 5, both peak
and cell-edge spectral efficiencies
degrade with the increase in the number
of floors because of more inter-floor
interference coordination effect, while
keeping max PM fixed. This is because,
as the number of floors increases, interfloor interference coordination effect
becomes significant so that the number of
resources reused to a specific femto user
(the maximum number of resources
reused to a femto user that define the
peak data rate) is decreased, hence peak
spectral efficiency. In addition, this in
turn also reduces the throughput per user,
i.e. cell-edge spectral efficiency.
However, this pitfall can be overcome
by increasing the value of max PM in the
MPF scheduler since increase in max PM
will allow the scheduler to reuse more
resources to a specific user even though
the number of femto users scheduled in a
TTI is the same.
On the contrary, with the increase in
max PM chosen in MPF scheduler, there
is a dramatic rise in both peak and celledge spectral efficiencies since more
reuse of resources to a specific femto user
is now possible, along with having an
opportunity of scheduling more femto
users for resource reuse, because of less
inter-floor interference coordination
effect. Hence, the effect of inter-floor
interference coordination, because of
increase in femtocell user density by
increasing the number of floors, can be
overcome by increasing the number of
163
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 149-168
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
Total users
Channel state information
Physical channels
Number of macro UEs per sector
Number of femto UEs per sector
Minimum distance between UE and macro BS1
Minimum distance from UE to femto BS1
UE speeds of interest
Carrier frequency band
Carrier aggregation
Aggregated carrier spectrum in LTE-Advanced
LTE carrier spectrum
Simulated bandwidth
Per RB bandwidth
Per CC RBs
RB reserved for macro users
Link considered1
Total macro BS transmit power1
Macrocell power control
Macro BS antenna gain, after cable loss1
Femto BS antenna gain1
Femtocell power control1
Noise
UE Noise Figure1
Path loss adjustment constant for 3.5 GHz (from
2 GHz)
Penetration loss (assume: UEs are indoors)1
Traffic model
Link level to system level mapping
Max PM (Mmax)
Fairness factor( tc)
TTI
Window size
Simulation runtime
1
taken from [28].
Value/Assumption
Dense Urban: macro layer has 7 sites (21 sectors)
with wrap-around, 1732 m Inter site distance (ISD),
and 20 % of macro UEs is outdoors.
1
1732 m
N (number of cells per row )
10
M (number of blocks per sector)
1
L (number of floors per block)
2
R (deployment ratio )
1.0
P (activation ratio)
100%
Probability of macro UE being
80%
indoors
100
Known
PDSCH (only data channel)
20
80
>= 35 m
20 cm
UE is stationary
3.5 GHz
Contiguous
80 MHz
20 MHz
20 MHz
180 kHz
100
6% (of total RBs)
Downlink
46 dBm
Fixed transmit power
14 dBi
5 dBi
Adaptive transmit power (0 to 20 dBm)
Thermal noise
9 dB
6 dB (additional)
10dB
Full buffer
MIESM
6
100 ms
1 ms
100 ms to 1000 ms
1000 ms
164
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 149-168
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
Table 4. Simulation scenarios for the evaluation of system performances.
Aspect
Scenario
objectives
Scenario 1
To investigate the effect of
interference coordination and
reuse of resources to femto
users.
User class
Traffic type
Femtocell
density
Simulation
level
Scenario
analysis
Scenario
findings
Scenario 3
To
investigate
the
performance of the MPF
scheduler for spectral and
energy efficiencies in LTEAdvanced systems.
Macro user and femto user
Full buffer
Fixed
Scenario 2
To
investigate
the
performance of the
MPF scheduler with
employing the proposed
new mechanism.
Macro user and femto
user
Full buffer
Fixed
System level
System level
1. Performance analysis of
the MPF scheduler as a
function of window size for
the following performance
metrics.
spectral efficiency
energy efficiency
Fairness performances
of macro users as well
as the system as a
whole.
Full buffer
Variable
165
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 149-168
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
Table 5. Average and peak spectral efficiency performances of the MPF and the PF Schedulers.
Efficiency
max
PM =
3,
floors
=2
Value
of
max
PM
and
no. of
floors
max
PM =
3,
floors
=6
max
PM =
7,
floors
=2
Table 6. Cell-edge and energy spectral efficiency performances of the MPF and the PF Schedulers.
Efficiency
Value
of
max
PM
and
Energy efficiency
max
PM =
3,
floors
=2
166
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 149-168
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
no. of
floors
max
PM =
3,
floors
=6
max
PM =
7,
floors
=2
2.
3.
4.
167
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 149-168
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
168
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 169-180
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
KEYWORDS
Cellular mobile communication, experiment,
laboratory, MATLAB, software tool.
1 INTRODUCTION
169
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 169-180
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
MATLAB software tools. With a
software based CMC laboratory,
students can model many features of
CMC, analyze and evaluate the
performance in both link and system
level. However, since a CMC laboratory
based on MATLAB tool is not obvious,
in this paper we present a novel
MATLAB based CMC laboratory.
The paper is organized as follows. In
section 2, the methodology used to
develop each experiment is discussed.
All experiments of the laboratory are
briefly described in section 3. The
laboratory assessment is incorporated in
section 4. We finish the paper with a
conclusion in section 5.
2 LABORATORY DEVELOPMENT
METHODOLOGY
In this section, we describe how the
development
of
each
laboratory
experiment is carried out. We consider a
stepwise approach to develop each
experiment
as
follows.
Step01:
introduction into which relevant
background,
problem
statement,
objective, and significance of the
experiment is described. Step02: system
or link model that incorporates three
parts: conceptual model, analytical
model, and simulation model. The
conceptual model incorporates the
system or link architecture and
configuration. The analytical model
incorporates necessary mathematical
expressions
that
transform
the
conceptual model into a methodical
demonstration. The simulation model
incorporates
typical
simulation
parameters, assumptions, and scenarios
which are used to simulate the system or
link behavior based on the analytical
model for performance evaluation. In
addition,
simulation
experimental
170
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 169-180
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
emphasize considerably on common
mathematical
functions,
complex
numbers, elementary matrices, vector
and matrix calculations, numerical
operations and transformations of
matrices, operator precedence, general
and logical functions, data manipulation
commands, and graphics. However,
because of MATLABs wide spread
applications, we limit our focus on using
those instructions that are more relevant
to simulate the objectives of this
laboratory experiments only.
3.2 Experiment 2: Propagation
Models and Path Loss Estimation in
Cellular Mobile Communication
In this experiment, we primarily carry
out the impact of carrier frequency (f)
and distance (d) on path loss. In
addition, a sensitivity analysis is carried
out that provides critical parameters such
as base station (BS) antenna height (hb),
mobile station (MS) antenna height (hm)
is incorporated for the system design and
planning purpose.
We consider the very optimistic Freespace model, the very pessimistic
International Telecommunication Union
Radiocommunication Sector (ITU-R)
model, and the more realistic Hata
model. All models are conceptually and
analytically described, followed by
respective
simulation
performance
evaluation. We finish this experiment
with a comparison of these path loss
model simulation results as shown in
figure 1.
) is
Eb
N0
Eb
N0
[5].
171
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 169-180
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
Hence, by plotting
Eb
N0
as a function
(a)
(b)
(c)
(d)
Figure 2. Frequency flat (a) Rayleigh fading and
(b) Rician fading channel responses and
frequency selective (c) Rayleigh fading and (d)
Rician fading channel responses.
172
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 169-180
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
In multipath channels, the transmitted
signal propagates over various paths, and
each path is characterized with relative
power and delay. The power-delay
profile of the channel is an indication of
the expected degree of dispersion of the
transmitted signal that defines the
maximum
possible
transmission
bandwidth (Bc) by evaluating delay
spread of the channel for proper
reception.
Similarly, the spectral broadening or
Doppler spread (fd) of the transmitted
signal, caused by the degree of relative
motion between BS and MS, defines the
maximum possible transmitted symbol
duration (Tc). Based on the relative
magnitude of Bc over Bs and Tc over Ts,
the channel can be classified as given in
table 1.
Table 1. Channel classification in mobile
communication.
Parameter
Delay
spread
Doppler
spread
Channel classification
Condition
Channel type
Frequency flat
Bc >> Bs
Frequency
Bc < Bs
selective
Slow fading
Tc >> Ts
Fast fading
Tc < Ts
No sectorization
1 8 15 22 29 36 43 50 57 64 71 78
85 92 99 106 113 120
2 9 16 23 30 37 44 51 58 65 72 79
86 93 100 107 114 121
3 10 17 24 31 38 45 52 59 66 73
80 87 94 101 108 115 122
4 11 18 25 32 39 46 53 60 67 74
81 88 95 102 109 116 123
173
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 169-180
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
E
F
G
Sector
index
A1
B1
C1
D1
E1
F1
G1
5 12 19 26 33 40 47 54 61 68 75
82 89 96 103 110 117 124
6 13 20 27 34 41 48 55 62 69 76
83 90 97 104 111 118 125
7 14 21 28 35 42 49 56 63 70 77
84 91 98 105 112 119 0
With sectorization
3-sectored cell
6-sectored
cell
1 22 43 64 85 106
1 43 85
2 23 44 65 86 107
2 44 86
3 24 45 66 87 108
3 45 87
4 25 46 67 88 109
4 46 88
5 26 47 68 89 110
5 47 89
6 27 48 69 90 111
6 48 90
7 28 49 70 91 112
7 49 91
Eb
, and it can be
N0
174
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 169-180
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
3.8 Experiment 8: Radio Resource
Allocations and Scheduling in Cellular
Mobile Communication
Radio resource allocation and
scheduling is one of the important areas
in mobile communication that needs
considerable attention to address several
significant issues such as spectral
efficiency, energy efficiency, and quality
of service. Scheduler defines the specific
usage of physical resources such as time,
frequency, and power. In this
experiment, students evaluate the
performance of the generic schedulers
such as Round Robin (RR), Max-SNR,
and Proportional Fair (PF) in terms of
average cell spectral efficiency and
Jains fairness index in a macrocell of
LTE-Advanced systems.
The system model considers a dense
urban deployment. A single macrocell is
considered. A number of macro users
(per cell for a 3-sector site) are assumed
within the macrocell coverage (figure 6).
Figure 6 shows the cell layout in the
system architecture, where macro users
are distributed randomly and uniformly
throughout the macrocell.
The throughput in any arbitrary
resource block (RB) can be expressed in
terms of the mutual information (MI) by
employing Shannons formula as given
below [8] [9].
0,
10dB
( dB)
4.4,
22dB
(2)
ISD
3R
A single macrocell
site
eff ( ) 10 log10 (2 1)
(3)
Hence, the system level spectrum
efficiency per RB per TTI can be
expressed as
(4)
Rsys = log 2 ( 1 + eff )
Jains fairness index is adopted for the
evaluation of the user fairness
performance to define the degree of
fairness among users. Mathematically,
the fairness index can be expressed as
follows [10].
2
uk
FJ k 1K
2
K uk
(5)
k 1
175
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 169-180
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
efficiency performance. On the contrary,
Max-SNR scheduler provides the best
spectral efficiency but the worst fairness
performances.
However,
the
PF
scheduler provides a trade-off between
fairness
and
spectral
efficiency
performances by taking the user
previous resource allocation history into
account.
10:
Design
of
176
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 169-180
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
Antenna
gain
Antenna
gain
Diameter
Fade margin
transmitter
NF
Cable
loss
Cable
loss
receiver
received
power
Transmit
power
A single-hop microwave link with no repeater
Distance
nu (t)
su (t)
Receiver
Down
converter
TWTA
Gu/Tu
EIRPs
su (t ) + n u (t )
s(t)
EIRP = Pt Gt
nu (t)
G/T
HPA
Pt
LNA
Gt
st nt
177
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 169-180
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
ratio, (C/N)total can also be estimated. If
(C/N)u >>(C/N)d, then C/N (C/N)d. In
this case the satellite link is said to be
downlink-limited.
When
(C/N)u
<<(C/N)d, then C/N (C/N)u, and the
satellite link is said to be uplink-limited.
We consider both single carrier and
multi-carrier per transponder for the link
budget estimation and carryout an
analysis of how multi-carrier scenario
affects the link performance. Followings
as given in table 2 are parameters that
are considered as an example scenario to
estimate the satellite link performance
for the operation of single carrier per
transponder.
Table 2. Default simulation parameters and
scenarios for satellite link [13].
Parameter
Transponder operation band
Noise bandwidth
Satellite antenna gain-to-noise
ratio
Satellite saturation EIRPs
Satellite TWTA BOi
Satellite TWTA BOo
Earth station transmit antenna
gain
Earth station receive antenna
gain
Earth station carrier power fed to
antenna
Maximum uplink and downlink
slant range
System noise temperature Ts
Uplink tracking loss
Downlink tracking loss
Interference into or from
adjacent satellite, terrestrial
interference, etc. and any
influence by the atmosphere
Value
14/12 GHz
46 MHz
-1.6dB/K
44 dBW
11 dB
6 dB
57.6 dB
56.3 dB
174 W
37,506 km
160 K
1.2 dB
0.9 dB
Negligible
Parameter
Carrier EIRP
Uplink free space loss
Uplink carrier-to-noise ratio
Satellite EIRP
Downlink free space loss
Downlink carrier-to-noise ratio
Total satellite link (earthsatellite-earth) carrier-to-noise
ratio
Satellite characteristics
Value
80.0055
dBW
206.8463 dB
23.3591 dB
44 dBW
205.5074 dB
24.8926 dB
127.2986 dB
uplink
limited
4 LABORATORY ASSESSMENT
In order to assess the significance of
the CMC laboratory, we carried out a
student survey within the registered
students of the CMC laboratory of fall
2012 semester at AIUB. The total
number of students participated in the
survey was ninety one. The survey
questions were categorized into two
parts: part I incorporates general
questions and part II incorporates query
based questions, with five questions in
each part (table 4). Students were asked
to put a tick on either Yes or No, with an
option for making comments, explaining
the reasons for their opinions.
Figure 11 shows the response of the
students for each question of the survey
(table 4). From figures 11 (a) and (b), it
can be found that overall most of the
students
Table 4. Student survey questions of CMC
laboratory assessment.
No
Q01
Q02
Q03
Question
Part I: General category
The laboratory experiments are relevant
and helpful for understanding the
contents of the course cellular mobile
communication.
The experiments of the lab are easy to
understand and have been written
properly with sufficient details.
The lab provides sufficient realization
of the practical aspects of cellular
178
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 169-180
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
Q04
Q05
Q06
Q07
Q08
Q09
Q10
(a)
(b)
Figure11. Response of the students to the
survey questions (a) general category (b) query
based.
5 CONCLUSION
In this paper, we introduced a
MATLAB based cellular mobile
communication
(CMC)
laboratory
course. The laboratory course comprises
of ten experiments, covering the
fundamental
design
parameters,
179
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 169-180
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
considerations, and estimations of CMC
both in the radio interface and core
network levels. A survey over ninety one
students was carried out to assess the
importance and impact of the laboratory.
Almost all students found the
laboratory experiments relevant and
helpful for understanding the CMC
theory course and for further research
and professional career development.
Since the laboratory is just MATLAB
software based, the laboratory can be
offered at affordable cost to provide
students to gain practical realization of
the CMC theory course concepts at the
university. The authors would be pleased
to share the laboratory course materials
with any individual or institution
interested in it and can be reached at
either of the authors email addresses.
5.
6.
7.
8.
9.
ACKNOWLEDGEMENT
The authors are thankful to Abdur
Rahman, Head of Undergraduate
Program and Farhadur Arifin, an
Assistant Professor of the EEE
department, AIUB, Bangladesh for their
generous support to develop the
laboratory.
10.
11.
REFERENCES
1.
2.
3.
4.
12.
13.
180
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 181-189
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
ABSTRACT
Wireless Sensor Network (WSN) is a
collection of small sensor nodes with
aptitude to sense, compute and transmit data
that are deployed to observe a physical
environment. The sensor node has limited
capabilities, especially its energy reserve, its
processing ability and its memory storage.
Data dissemination and gathering protocols
design for WSN are crucial challenges since
those protocols should be easy, energyefficient, and robust to deal with a very large
number of nodes. Also, they should be selfconfigurable to node failures and dynamic
changes of the network topology. In this
paper, we present a new algorithm for
gathering sensor reading based on chain
forming using Ant Colony Optimization
(ACO) technique. To allow network lifetime
extension, the ACO provides the shortest
network nodes chaining instead of starting
from the furthest node and using Greedy
algorithm as PEGASIS do. The leader role
duration is defined for each node based on
its required energy to do this role in the
established chain. Which avoids fast nodes
energy depletion and then, the network
lifetime would be extended. Through
simulation, it is proved that the proposed
algorithm allows network stability extension
compared to the most known chaining
algorithm.
KEYWORDS
1 INTRODUCTION
A Wireless Sensor Network (WSN) is
composed of a collection of tiny and
lightweight sensor nodes deployed in
large number to monitor the surrounding
conditions [1]. Those nodes cooperate to
collect, process and transmit the
surrounding
conditions
such
as
temperature, humidity, vibration The
WSN have many application domains
such as environmental survey, smart
home,
medical
and
agricultural
monitoring, etc.
Some of the early literature works on
WSNs have discussed the benefits of this
kind of networks compared to MANETs
[2, 3, 4]. WSNs have main advantages
over the conventional networks deployed
for the same purpose such as greater
coverage, accuracy, reliability and all of
the above at a possibly lower cost.
Since the sensor node has small size, and
maybe deployed in hostile region, its
available energy is considered as the
main constraint. Therefore, much
attention must be given to the energy
consumption when designing protocols
for this kind of networks [5, 6].
WSN research domain is much active so,
in the last few years, a variety of
protocols have been proposed for
prolonging the WSN duration service
181
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 181-189
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
when gathering the sensor reading to the
sink. The most of those protocols can be
divided as either flat or hierarchical
based.
In densely deployed sensor network, an
event is often detected by more than one
sensor and duplicated data are generated.
This data redundancy is often
eliminated, which cannot only decrease
the global data to be transmitted and
localized most traffic within individual
groups, but also, reduces the traffic and
consequently, contention in a wireless
network, which allows decreasing
energy consumption.
Data aggregation is a way to reduce
energy consumption. This aggregation
consists of suppressing redundancy in
different data messages. This is the key
idea for the most hierarchical routing
protocols.
In addition, scalability is one of the key
design attributes of sensor networks.
Since a single-tier network can conduct
the gateway to overload with the
increase in sensors density, the main
objective of hierarchical routing is to
efficiently
preserve
the
energy
consumption of sensor nodes by
involving
them
in
multi-hop
communication within a particular
cluster and by performing data
aggregation and fusion that decrease the
number of data messages that would be
transmitted to the sink.
In literature, many research works have
explored hierarchical routing in WSN
from different perspectives. Some of the
hierarchical protocols are LEACH [7, 8],
PEGASIS [9], TEEN [10,11], SEP [12],
DEEC [13] and APTEEN [14].
LEACH is the most popular energyefficient
hierarchical
clustering
algorithm that was proposed for
reducing power consumption in WSN by
doing data aggregation. It uses clustering
182
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 181-189
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
because it requires only one designated
node to send one combined data to the
BS.
A greedy algorithm is proposed to be
used in PEGASIS to form the chain;
each node selects the closest neighbor
that is not chained yet and so as until all
network nodes are chained. This
algorithm is executed before starting the
first round of data transmission. To
construct the chain, PEGASIS starts with
the furthest node from the BS. This is to
make sure that nodes farther from the BS
have close neighbors, since in the greedy
algorithm the neighbor distances will
augment gradually because nodes
already on the chain cannot be revisited
again. After chain formation, data
transmission phase will start. In this
phase, node can deplete its residual
energy, and then the chain will be
reconstructed in the same manner to
avoid the dead node.
183
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 181-189
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
PEGASIS [9] and Seetharam et al. [15]
propose that the chain forming starts
from the farthest node to the BS. As the
chain length will not be the shortest one,
the network lifetime is not optimized.
(1)
E RX l.E elec
(2)
184
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 181-189
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
the edge (i,j) is updated as follow:
m
(5)
ij
ij .(1 ) ijk
k
Otherwise
185
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 181-189
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
different energy when being leader
depending on its location in the chain
and its distance to the BS, which means
that a certain nodes would die quickly.
So, let network node plays leader
depending on its energy capacity can
ensure best energy load balancing.
Let consider a network formed by N
chained nodes and Ei0 is the initial
battery energy of the node i. Let EiBS is
the required energy for the node i to
transmit to the base station, Eij is the
energy to transmit from the node i to
node j and Erec is the energy cost
required for receiving a packet.
The node i deployed in the monitored
area can do Ti transmissions in its
lifetime (until the depletion of its
residual energy) in which it is the leader.
For notation simplicity, we suppose that
the chain nodes are organized as follow;
{1, 2...N} and then the end chain nodes
are 1 and N. For the node 1, that will be
leader for T1 times, will do T1
transmission
to
the
BS
and
T2+T3+...+TN transmissions to the next
node in the chain, lets node 2. Then, the
equation (7) must be respected.
E1BS E rec
E12
...
E12 T1 E10
... T2 E 20
E 21 E rec E 2BS 2.E rec E 23 E rec
.
(8)
...
...
...
... ...
E 21
...
E NBS E rec TN E N 0
E 21
186
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 181-189
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
which do data aggregation and so that
until reaching the leader. This leader
aggregates the received data with its own
and transmits it the sink.
Since the leader consumes considerable
energy, and in order to distribute energy
load, each node i that becomes leader
plays this role for Ti consecutive
transmissions rounds.
4 SIMULATION PARAMETERS
AND RESULTS
To evaluate the performance of our
technique, several MATLAB simulations
were performed and the represented
results are an average. We consider a
square network with N nodes deployed
randomly in the field. The used
parameter values in our work are given
in Table1.
Table1: Simulation Parameter Values.
Description
Symbol
Network dimension
Xm*Ym
Number of network N
nodes
Data packet length
L
Electronic Energy
Eelec
Amplifier Energy
Eamp
Value
100m*100m
10-200
2500 Bits
50nJ/bit
100pJ/bit.m2
187
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 181-189
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
The simulation results presented above
illustrate that the proposed scheme
exploits judiciously the network energy
since it selects the short network chain
so that the transmission energy is
optimized. It exploits, also, the capacity
to play leader role for each node to
ensure a best energy load repartition
over chain nodes.
5 CONCLUSION
Network nodes chaining is a better solution
to extend network lifetime since only the
leader node that do long transmission to the
BS. In this paper, we proposed an Ant
colony optimizing technique to form
network nodes chain ensuring that this chain
is the short one. Chain nodes become leader
for a lot of time based on its required energy
to transmit to the BS and to its two chain
neighbors. The proposed technique allows
extra network transmission rounds as
compared to PEGASIS. This extension is
guaranteed for different number of network
nodes, for different BS location and for
different nodes initial energies. The
simulation results show that our technique
outperforms to the well-known protocol for
chaining in wireless sensor networks
(PEGASIS protocol). In future, we will
continue the work investigating the effect of
data correlation on the network performance
for this kind of routing protocols.
6 REFERENCES
1.
2.
3.
188
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 181-189
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
13. Wang M.W., Qing L., Zhu Q.X.: Design of
a distributed energy-efficient clustering
algorithm for heterogeneous wireless sensor
networks.
Computer
Communication
(Elsevier), 29(12): 2230-2237, 2006.
14. Manjeshwar A., Agrawal D. P.: APTEEN: A
Hybrid Protocol for Efficient Routing and
Comprehensive Information Retrieval in
Wireless Sensor Networks. In: Proc. 2nd
International Workshop on Parallel and
Distributed Computing Issues in Wireless
Networks and Mobile computing, San
Francisco CA, April 2001, pp. 2009-1015.
15. Seetharam A., Acharya A., Bhattacharyya
A., Naskar M. K.: An Energy Efficient Data
Gathering Protocol for Wireless Sensor
Networks. Journal of Applied Computer
Science, no. 1(2). 2008 pp. 30-34.
16. Dorigo, M., Maniezzo, V. and Colorni, A.:
The ant system: Optimization by a colony of
cooperating agents, IEEE Transaction on
System Man, and Cybernetics-Part B
26(1996), pp. 2941.
189
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 190-196
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
ABSTRACT
Vehicular Ad-hoc network is an emerging
class of wireless network that provide
efficient communication between vehicles
and between vehicles and road side units.
VANETs are gaining attention from both
industrial and academic point of view due to
its applications ranging from safety
applications to infotainment applications.
The main component for success of VANET
applications is its routing. Design of an
efficient routing protocol is important for
smart Intelligent Transport System (ITS). In
this paper, we consider how traffic lights
affect the design of routing protocol in urban
vehicular
communication.
In
urban
environment the roads are partitioned to
several segments due to the traffic lights.
The design of a routing protocol in such an
environment is difficult. This paper proposes
a new intersection based routing protocol
which takes traffic lights in to consideration
and with low overhead.
KEYWORDS
VANET, ITS, Intersection based routing,
broadcast storm problem.
1 INTRODUCTION
During the last few years vehicular
communication is attracting growing
attention from both academic and
industrial point of view. This is because
190
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 190-196
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
such as Ad-hoc On Demand Distance
Vector routing (AODV) [2], Dynamic
Source Routing [3] etc. But these
protocols fail to address certain peculiar
characteristics of VANET. Various
routing protocols especially for VANET
were proposed to make routing more
efficient and reliable.
In this article we consider the vehicular
communication in highly dense urban
environment. Intersection based routing
protocols are highly reliable in such
environment. In intersection based
routing, when vehicles move on straight
road, they forward by greedy
forwarding. When they reach an
intersection a decision is made whether
to forward in same direction or to
perpendicular direction.
Many intersection based routing
protocols have been proposed to carry
efficient routing in VANET. But only
few protocols consider traffic lights. The
existing protocols that consider traffic
lights are having high overhead. This
paper proposes a new routing protocol
having low overhead. The remaining of
this paper is organized in to following
sections: Section II gives related works,
Section III describe the problem with
traffic lights, Section IV gives the new
approach and finally Section V
concludes the paper.
2 RELATED WORKS
VANET routing protocols can be
categorized in to two: topology based
and geographic (position based) routing
protocols. In topology based routing
protocols each node is expected to know
the entire network topology. In
Geographic or position based routing
protocols the decision on routing is
based on the position of the sender,
position of the destination and position
of the senders one hop neighbors using
GPS. It is assumed that each node knows
its position and the position of the
destination. The position of its one hop
neighbors is obtained from periodically
exchanged beacons. In geographic
routing the messages can be forwarded
to destination without knowing the
topology and without prior route
discovery. This section briefly describes
some prominent geographic routing
protocols.
2.1 Greedy Perimeter Coordinator
Routing (GPCR)
Greedy Perimeter Coordinator Routing
[4] makes use of street and junctions to
forward packets. GPCR consists of 2
parts: a restricted greedy forwarding
procedure and a repair strategy.
In all intersection based routing
protocols the actual routing decision is
made on junctions. In GPCR nodes on
junctions are called coordinators.
Packets are always forwarded to these
coordinators rather than to nodes across
the junction. If coordinators are not
present, the packets are forwarded to
node with the largest distance from the
forwarding node. If more than one
coordinator is present then one is taken
randomly and packet is forwarded to this
191
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 190-196
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
coordinator. In this protocol a decision is
made greedily so that a node with great
progress to destination is always
selected. When packet reaches a local
optimum, Right Hand Rule is used to see
to which street the packet should follow
as the repair strategy.
GPCR is effective as it does not need
any external information and graph
planarization algorithms. But first part of
GPCR fails on curve road and second
approach fails on sparse road. It selects
street without understanding whether
there are sufficient nodes to carry the
packet.
2.2 Improved Greedy Traffic Aware
Routing (GyTAR)
GYTAR [5] is an intersection based
routing protocol. It consists of 2
modules: dynamic junction selection
through which a packet must reach
destination and an improved greedy
strategy used for forwarding. A
forwarding node looks for the
neighboring junctions in digital map.
Then it assigns a score to each of these
junctions considering the traffic density
and curve metric distance to destination.
The score is calculated as in Eq. 1.
ScoreJ=[1-Dp]+[min(Navg/Ncon,1)] (1)
192
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 190-196
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
VADD defines three packet models:
Straight
road,
Intersection
and
Destination.
(2)
193
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 190-196
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
geographical location of vertex acts as a
vertex. This process continues until the
packet reaches the proximity of last
vertex. When the last vertex proximity is
reached the packet is greedily forwarded
to the destination.
Disconnected link
194
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 190-196
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
4 RED LIGHT FIRST
FORWARDING
Shortest Path Based Traffic Aware
Routing (STAR) protocol is an
intersection based routing protocol that
takes traffic light in to consideration.
STAR also makes use of the vehicles
turning right from green light segment to
red light segment to ensure end to end
connectivity on a red light segment. In
STAR the segment connectivity is
measured by broadcasting connectivity
information. This broadcast mechanism
leads to serious problems like
redundancy, contention and collision of
messages. This problem is referred as
broadcast storm problem [11]. Thus
STAR protocol introduces much
overhead in the network.
In this paper we tackle this problem by
introducing a new intersection based
routing protocol with low overhead. This
paper proposes a protocol Red Light
First Forwarding (RLFF) which avoids
blind flooding to measure connectivity.
In Red Light First Forwarding, it is
assumed that each vehicle is employed
with GPS and each node knows its
position, position of its neighbours and
position of destination.
On a straight road, the vehicles forward
packets by greedy forwarding together
with carry and forward recovery
mechanism. In greedy forwarding when
a node wants to send a message to
destination, it checks whether the
destination is reachable. If the
destination is in its transmission range
the message is forwarded directly. If
destination is not reachable, it forwards
the packet to an intermediate node which
is near to the destination than itself.
When it reaches an intersection, it
checks whether the red light segment
195
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 190-196
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
If dmin < D go to step 4. Otherwise
resume the interrupted waiting.
Step 4: Cancel retransmission. Exit the
procedure.
RLFF is designed to have low network
overhead, minimum delay and high
delivery ratio.
5 CONCLUSION
VANET have recently been the topic of
extensive research due to its wide range
of application. Routing is the main
component for success of VANET
applications. There are lots of routing
protocols designed especially for
VANET. But most promising for urban
scenario is intersection based routing
protocols. This paper deals with an
intersection based routing protocol that
consider traffic lights and having low
overhead. Designing of routing protocols
that handle all applications is not at all
practical.
This approach will be simulated in ns2
[12] by generating mobility models
using
VanetMobiSim
[13].
The
performance of new protocol will be
measured in terms of network overhead,
delivery ratio and delay and will be
compared with existing protocols like
STAR and GyTAR.
6 REFERENCES
1. Yasser Toor, Paul Muhlethaler, Anis
Laoutti and Arnaud De La Fortelle
Vehicular
Ad-hoc
Networks:
Applications and related technical
issues,
IEEE
Communications
Surveys, 3rd Quarter 2008, Vol 10,
No.3.
2. Perkins, C.; Belding Royer, E.; Das,
SAd-hoc On Demand Distance Vector
Routing, IETF RFC, 3561, July 2003.
196
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 197-205
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
1 INTRODUCTION
Normalization is, in relational database
design, the process of organizing data to
minimize redundancy. It usually
involves dividing a database into two or
more tables and defining relationships
between the tables. The objective is to
minimize modification problems that
could arise after modification of a table
field. Edgar Codd, the inventor of the
relational model, also introduced the
197
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 197-205
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
F={ABCDEF; BCADEF; B D;
EF }.
2 BASIC DEFINITIONS
198
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 197-205
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
removing
redundant dependencies,
called minimal cover [9], [10]. To
achieve this goal, one needs to compute
the closure of a given set of attributes
and then remove redundant attributes.
Let X be a set attributes and F be a set of
functional dependencies. Let X+ be the
set of all attributes that depend on a
subset of X with respect to F, i.e., X+ is
the set of attributes Z, such that XZ
F. X+ is called the closure of X w.r.t F.
The Algorithm 1 computes the closure of
a given set of attributes w.r.t F:
Algorithm 1: computes X+.
Input: A relation R, a set of functional
dependencies F and a set X of attributes.
Output: X+, the closure of X.
X+:= X
While there is a fd Y A F
If YX+ and A X+ then
X+:= X+ A
End if
End while
199
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 197-205
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
Example 1:
Suppose we want to normalize the 1NF
relation ClientRental up to 3NF. This
example is taken from [12] with minor
modifications.
This relation contains 9 attributes and
18 dependencies:
Clientrental(clientNo, propertyNo,
clientName, pAddress, rentStart,
rentFinish, rent, ownerNo,
ownerName)
with the following dependencies F:
f 1:
f 2:
f 3:
f 4:
f 5:
f 6:
f 7:
f 8:
f 9:
f10:
f11:
f12:
f13:
f14:
f15:
f16:
f17:
f18
propertyNo,rentStart,
ownerNo
clNo,propertyNo,
clName
clNo,rentStart,
clName
clNo
propertyNo,rentStart
clNo,propertyNo
ownerNo
propertyNo
clNo,rentStart
clNo,rentStart
propertyNo
propertyNo
clNo,rentStart
propertyNo
clNo,rentStart
clNo,rentStart
propertyNo,rentStart
propertyNo, clNo
rentFinish
rentFinish
rentFinish
clName
clName
rentStart
oName
oName
oName
pAddress
pAddress
rent
rent
ownerNo
ownerNo
propertyNo
clNo
ownerNo
Initialization: F':=F
For the fd f1, we have
+
rentFinish (propertyNo,rentStart) ;
propertyNo,rentStart
clNo,propertyNo
clNo,rentStart
clNo
propertyNo,rentStart
clNo,propertyNo
ownerNo
propertyNo
clNo,rentStart
clNo,rentStart
propertyNo
propertyNo
clNo,rentStart
propertyNo
clNo,rentStart
clNo,rentStart
propertyNo,rentStart
propertyNo, clNo
rentFinish
rentFinish
rentFinish
clName
clName
rentStart
oName
oName
oName
pAddress
pAddress
rent
rent
ownerNo
ownerNo
propertyNo
clNo
ownerNo
200
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 197-205
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
clNo,propertyNo
clNo
clNo,propertyNo
ownerNo
propertyNo
propertyNo
propertyNo
clNo,rentStart
propertyNo,rentStart
rentFinish
clName
rentStart
oName
pAddress
rent
ownerNo
propertyNo
clNo
201
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 197-205
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
clNo,propertyNo
clNo
clNo,propertyNo
ownerNo
propertyNo
propertyNo
rentFinish
clName
rentStart
oName
pAddress
rent
f14:
f16:
f17:
propertyNo
clNo,rentStart
propertyNo,rentStart
ownerNo
propertyNo
clNo
5 NORMALIZATION
5.1 The Second Normal Form
A relation is in second normal form
(abbreviated 2NF) if it is in 1NF and no
202
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 197-205
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
203
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 197-205
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
clNo,propertyNo
clNo
clNo,propertyNo
ownerNo
propertyNo
propertyNo
propertyNo
clNo,rentStart
propertyNo,rentStart
rentFinish
clName
rentStart
oName
pAddress
rent
ownerNo
propertyNo
clNo
}
We can say that the set transitive
dependencies is
Ft={ ownerNooName } as ownerNo
is not a subset of a candidate key and
oName is not a key-attribute.
204
International Journal of Digital Information and Wireless Communications (IJDIWC) 3(2): 197-205
The Society of Digital Information and Wireless Communications, 2013 (ISSN: 2225-658X)
6 CONCLUSION
In this paper we have presented
algorithms for relational database
normalization into 2NF and 3NF using
their general definitions in a step-by-step
feature. The first step before performing
the procedure is to make a preprocessing
on the set of dependencies to remove
redundant dependencies. We have tested
our algorithms on many realistic
examples with multiple candidate keys
taken from different sources.
This work has mainly the following
major advantages: (i) the general and
original definition of normal forms is
used, (ii) the removal of redundant
dependencies, (iii) in all phases, the
computation of attributes closure are
minimized compared to other algorithms
although using a restricted definition of
normal forms, and (iv) a primary key is
determined for any generated relation.
8 REFERENCES
1.
205