Content uploaded by Dino Golić
Author content
All content in this area was uploaded by Dino Golić on Aug 30, 2018
Content may be subject to copyright.
17th International Symposium INFOTEH-JAHORINA, 21-23 March 2018
978-1-5386-4907-7/18/$31.00 ©2018 IEEE
Impact of Mobile Ad-hoc Network Parameters on
Quality of Experience for Video Streaming
Dino Golić
Faculty of Electrical Engineering
University of Sarajevo
Sarajevo, B&H
dino.golic13@gmail.com
Jasmina Baraković Husić
Faculty of Electrical Engineering
University of Sarajevo
Sarajevo, B&H
jasmina.barakovic@etf.unsa.ba
Sabina Baraković
Ministry of Security of
Bosnia and Herzegovina
Sarajevo, B&H
barakovic.sabina@gmail.com
Abstract—Video streaming over MANETs (Mobile Ad-hoc
Networks) is challenging due to specific features that characterize
this network. The importance of MANET-specific features has
been already emphasized in terms of QoS (Quality of Service).
Nevertheless, these parameters have been considered to a limited
extent in terms of QoE (Quality of Experience), and thereby, they
still remain not well understood. Therefore, this research has been
motivated by the challenge to provide a deeper understanding of
MANET-specific parameters and their impact on QoE for video
streaming. Experimental study has included two phases, i.e., video
streaming emulation over MANET and end-user survey to collect
MOS (Mean Opinion Score) rates. Results obtained from
statistical analysis imply that there exists strong and statistically
significant impact of individual MANET-specific parameters and
their interaction on QoE for video streaming.
Keywords-ANOVA; MOS; MANET; QoE; video streaming
I. INTRODUCTION
The 5G (5th Generation) mobile networks are proposed to
offer almost infinite networking capability and unrivalled user
experience for mobile users at any time and in any place or
situation [1]. 5G networks involve and benefit from many
technical advances including SDN (Software Defined
Networking), NFV (Network Functions Virtualization),
MANETs (Mobile Ad-hoc Networks), data mining, etc.
MANETs have a specific purpose and infrastructure. Since
they are not based on the existing infrastructure such as access
points or routers, each node participates in data forwarding
[2]. MANETs are commonly used in areas where standard
infrastructure is difficult to develop. Therefore, they can play
an imperative role in building up a strong foundation for 5G
networks [3]. Giving the fact that video produces generally the
most demanding type of traffic in terms of network resources,
realizing video streaming over MANETs is quite challenging.
Moreover, MANETs are characterized by resource constraints,
high dynamics, and frequent transmission errors [4], which
can affect the user experience and its quality in the context of
video streaming.
In order to improve user experience and offer maximized
quality, a deep understanding of the IFs (Influence Factors)
that affect QoE (Quality of Experience) is needed [5]. A QoE
IF has been defined as “any characteristic of a user, system,
service, application, or context whose actual state or setting
may have influence on the quality of experience for the user”
[6]. Although there are several approaches to the
categorization of QoE IFs [7]-[9], we consider the
classification proposed by EU Qualinet community which
groups them into following three categories [6]: (1) HIFs
(Human IFs); (2) SIFs (System IFs); and (3) CIFs (Context
IFs). In this paper, the focus is on the SIFs which refer to the
features and characteristics that determine the technical quality
of the application or service. Furthermore, SIFs can be divided
into four categories: media-related, content-related, network-
related, and device-related SIFs. Here, we focus on network-
related SIFs referring to data transmission over MANETs, or
more precisely, on specific parameters that characterize
MANETs (i.e., routing protocol type, network size, number of
nodes, node mobility, node speed, etc.).
The importance of MANET-specific parameters has been
already emphasized in terms of QoS (Quality of Service) [10]-
[21]. However, in previous research studies on MANET-
specific parameters, their impact on QoE has been analyzed to a
limited extent [22]-[26], and, thereby, it still remains not well
understood. Therefore, this paper has been motivated by the
challenge to provider a deeper and more comprehensive
understanding of MANET-specific parameters and their impact
on QoE in the context of video streaming. In this regard, the
objective of this paper is to investigate the impact of MANET-
specific parameters (i.e., routing protocol type, number of
nodes, and speed of nodes) and their interactions on QoE in
terms of MOS (Mean Opinion Score) for video streaming. In
order to accomplish this objective, we have performed
experimental study that includes end-user survey to collect
MOS rates for video streaming over MANET, which are further
statistically analyzed by ANOVA (Analysis of Variance).
The rest of the paper is organized as follows. Section II
provides the non-exhaustive review of research activities
considering MANET-specific parameters and their impact on
QoE for video streaming. Section III describes the
experimental study which includes two phases, i.e., video
streaming emulation over MANET and end user survey for
collecting MOS rates. Experimental results obtained from
statistical analysis are presented and discussed in Section IV.
Finally, Section V summarizes the conclusions and presents
open issues for further research.
978-1-5386-4907-7/18/$31.00 ©2018 IEEE
II. RELATED WORK
This section provides the non-exhaustive review of
research activities considering the MANET-specific
parameters and their impact on QoE in the context of video
streaming. As mentioned in the Introduction, QoE is affected
among others by network-related SIFs, which refer to data
transmission over a network, i.e. delay, jitter, loss, throughput,
etc. There are many related works considering how these
factors depend on the MANET-specific parameters [10], such
as routing protocol, network size, node mobility or speed, etc.
A number of routing protocols can be used in MANETs,
but not all of them are suitable for video streaming. A. B.
Malany et al. [11] has shown that STAR (Source Three
Adaptive Routing) and RIP (Routing Information Protocol) do
not provide good performance in terms of throughput, so they
are not very convenient for video streaming over MANET. In
this regard, higher throughput is achieved by using Bellman-
Ford routing protocol. Moreover, S. Hakak et al. [12] [13] has
analyzed the impact of routing protocols along with packet
size and DSSS (Direct Sequence Spread Spectrum) rate on
delay and jitter. It has been found that AODV (Ad hoc On-
Demand Distance Vector) protocol achieves better
performance with DSSS rate below 2 Mbps, whereas DYMO
(Dynamic Manet on Demand) protocol realizes better
performance with DSSS rate reaching 5.5 Mbps. In addition,
S. S. Srivastava et al. [14] has discussed the impact of static
and dynamic routing protocols and proposed a new routing
protocol for minimalizing packet losses in MANETs.
To evaluate the performance of routing protocols in terms
of throughput, delay, and jitter, M. K. Gulati et al. [15] has
performed a simulation study by varying the network mobility,
size, and load. It has been found that DSDV (Destination
Sequenced Distance Vector) protocol is suitable for limited
network size and low mobility, DSR (Dynamic Source
Routing) protocol is preferable for moderate mobility, and
AODV protocol performs better among the previous two for
large network size and high mobility. The impact of different
mobility models (i.e., RWP (Random Way Point), RD
(Random Direction), and Mbg-SS (Mobgen Steady State)) on
throughput and delay has been studied in more detail by M.
Amnai et al. [16]. Additionally, M. Kumar et al. [17] has
performed another research study comparing the throughput,
delay, and jitter in MANETs with different routing protocols
and found that DSR protocol performs better compared to
other routing protocols. Similar results were obtained by S.
Barakovic et al. [18] in simulation study comparing AODV,
DSR, and DSDV routing protocols. On the other side, R. K.
Jha et al. [19] has shown that AODV is superior than DSDV
and OLSR (Optimized Link State Routing) in terms of
throughput, whereas OLSR gives better results than AODV
and DSDV in terms of delay, jitter, and loss.
MANET performance can be improved with the
transmission power increase [20]. Therefore, T. R. Sheltami et
al. [21] has analyzed the performance of WEAC/VBS-O
(Warning Energy Aware Clusterhead/ Virtual Base Station-On
demand) protocol in terms of power minimization aspect, and
additionally found that 2 Mpbs is not suitable for video traffic.
Instead 5.5 Mbps link speed is necessary for acceptable
performance.
Furthermore, there are some related works considering the
impact of these MANET-specific parameters on QoE for
video streaming. For example, P. T. A. Quang et al. [22] has
discussed a QoE-based routing solution by using PSQA
(Pseudo-Subjective Quality Assessment) tool that evaluate the
QoE in terms of MOS. However, there are many more related
works considering the impact of different routing protocols on
QoE for VoIP (Voice over IP) such as those performed by M. S.
Islam et al. [23], P. Sondi et al. [24], S. Mahajan et al. [25], and
V. Goyal et al. [26].
Bearing the aforementioned discussion in mind, one may
conclude that the influence of MANET-specific parameters on
QoE for video streaming has been studied not nearly as much
as their impact on QoS (i.e., network-related SIFs).
Summarizing related work presented in Table I, one can
conclude that the most analyzed network-related SIFs are
delay, jitter, throughput, and loss. Based on the literature
review, it can be noticed that these network-related SIFs are
commonly affected by routing protocols. Moreover, Table I
shows that the number of nodes followed by their speed are
typically varied parameters in most simulation studies.
Therefore, this paper aims to investigate the impact of routing
protocol type, number of nodes, and their speed on user
experience and its quality for video streaming over MANETs.
III. RESEARCH METHODOLOGY
After conducting the literature review [10]-[26], we have
concluded that it would be beneficial to investigate the impact
of routing protocol, number of nodes, and their speed on QoE,
which is collected in terms of MOS rates. That is why we have
formulated the following null hypotheses and conducted the
statistical analysis in order to test them.
H1: The differences in QoE caused by the routing protocol
impact in MANETs are not statistically significant.
H2: The differences in QoE caused by the number of
nodes impact in MANETs are not statistically significant.
H3: The differences in QoE caused by the speed of nodes
impact in MANETs are not statistically significant.
H4: The differences in QoE caused by the impact of
interaction between the routing protocol and number of nodes
in MANETs are not statistically significant.
H5: The differences in QoE caused by the impact of
interaction between the routing protocol and speed of nodes in
MANETs are not statistically significant.
H6: The differences in QoE caused by the impact of
interaction between the number of nodes and their speed in
MANETs are not statistically significant.
H7: The differences in QoE caused by the impact of
interaction between the routing protocol, number of nodes,
and their speed in MANETs are not statistically significant.
The abovementioned hypotheses are formed in the way that
we do not expect the existence of statistically significant impact
according to the rules of ANOVA statistical test (p>0.05).
978-1-5386-4907-7/18/$31.00 ©2018 IEEE
TABLE I. REVIEW OF RELATED WORKS CONSIDERING THE IMPACT OF MANET-SPECIFIC PARAMETERS ON NETWORK-RELATED SIF
Authors
[Number of References]
Network-
related SIF
Simulator
Routing
Protocol
Number of
Nodes
Node Speed
[m/s]
Mobility Model
Network Size
[m×m]
D.D. Perkins et al. [10]
Throughput
GloMoSim
AODV
DSR
50, 80
5, 40
RWM
1600×400
A.B.Malany et al. [11]
Delay
Throughput
QualNet
AODV
Fisheye
DYMO
STAR
RIP
Bellman Ford
LANMAR
LAR
2, 10, 50, 100
[0, 10]
RWM
1500×1500
S. Hakak et al. [12]
Jitter
QualNet
DYMO
AODV
100
30
RWM
1500×1500
S. Hakak et al. [13]
Delay
Jitter
QualNet
DYMO
AODV
100
30
RWM
1500×1500
M. K. Gulati et al. [15]
Delay
Jitter
Throughput
ns-2
DSDV
AODV
DSR
50
[0, 20]
RWM
1000×1000
M. Amnai et al. [16]
Delay
Throughput
ns-2
AODV
10, 20, 30, 40,
50, 60, 70, 80,
90, 100
10
Mbg-SS
RD
RWM
1000×1000
M. Kumar et al. [17]
Delay
Jitter
QualNet
ANODR
AODV
DSR
DYMO
LANMAR
IERP
ZRP
5, 10, 15, 20
NA
NA
100×100
S. Barakovic et al. [18]
Delay
ns-2
AODV
DSR
DSDV
50
20
RWM
500×500
R. K. Jha et al. [19]
Delay
Jitter
Loss
Throughput
ns-3
AODV
DSDV
OLSR
5, 10, 15
NA
NA
NA
S. Shrivastava et al. [20]
Delay
Jitter
Throughput
QualNet
AODV
50
NA
RWM
1500×1500
T. R. Sheltami et al. [21]
Delay
Throughput
NA
WEAC/VBS-O
NA
1.4
NA
2000×2000
Legend: ANODR (ANonymous On Demand Routing); AODV (Ad hoc On-Demand Distance Vector); DSDV (Destination Sequenced Distance Vector); DSR (Dynamic Source Routing); DYMO (Dynamic Manet on
Demand); GloMoSim (Global Mobile System Simulator); IERP (Inter-zone Routing Protocol); LANMAR (Landmark Routing Protocol); LAR (Location-Aided Routing); MbgSS (Mobgen Steady State); NA (Not
Applicable); OLSR (Optimized Link State Routing); RD (Random Direction); RIP (Routing Information Protocol); RWM (Random Waypoint Mobility); STAR (Source Three Adaptive Routing); VBS-O (Virtual
Base Station On-demand); ZRP (Zone Routing Protocol); WEAC (Warning Energy Aware Clusterhead).
However, it is expected to have statistically significant impacts
of individual parameters and their interactions on QoE (p<0.05)
because these MANET-specific parameters affect the network-
related SIFs as stated in literature review [10]-[21], and
consequently influence the QoE. In order to obtain the data
necessary to investigate the impact of MANET-specific
parameters (i.e., routing protocol type, number of nodes, and
speed of nodes) on QoE, experimental study has been
performed. The experimental procedure included two phases:
(1) video streaming emulation over MANET, and (2) end-user
survey with the aim of collecting MOS rates.
A. Phase 1: Video Streaming Emulation over MANET
The first phase involves two steps: (1) preparing the video
sequences, and (2) defining MANET configuration parameters.
To prepare the video sequence, we have used Cable Car stock
[27] in .mp4 format, which is characterized by duration of 39
seconds and resolution of 1920×1080 pixels. This video sequence
is shortened to 20 seconds along with audio content removal by
using ffmpeg tool [28]. As such, it was used for video streaming
emulation over MANET, which has been performed in network
simulator ns-3 and using EvalVid tool [29].
A total of 56 test scenarios have been created based on the
variation of the MANET configuration parameters that are
shown in Table II. Variation values are selected according to
configuration parameters setup used in related works as
summarized in Table I. Test scenarios differ from each other by
routing protocol (AODV, DSDV), number of nodes (20, 30, 35,
40), and their speed (1, 5, 7, 10, 12, 15, 30 m/s). They have been
used to create the same number of video sequences which are
used in the second phase to collect the MOS rates.
To investigate how MANET-specific parameters affect
QoE, impact of transmitter and receiver motion has been
examined. Each video sequence is sent from one transmitting
node to the receiver along with UDP (User Datagram Protocol)
background traffic. RWM (Random Waypoint Model) has been
our preferred mobility model since it is used for describing
random node motion in 2D (Two Dimensional) space. The
starting distance between nodes has been 15 meters.
978-1-5386-4907-7/18/$31.00 ©2018 IEEE
TABLE II. MANET CONFIGURATION PARAMETERS
Configuration parameters
Variation values
Routing protocol
AODV, DSDV
Number of nodes
20 ,30, 35, 40
Node speed [m/s]
1,5,7,10,12,15,30
MAC type
802.11
Transmission technology
OFDM 6Mbps
Propagation model
Friis
Duration of simulation [s]
30
Legend: AODV (Ad hoc On-Demand Distance Vector); DSDV (Destination Sequenced Distance
Vector); MAC (Media Access Control); OFDM (Orthogonal Frequency-Division Multiplexing).
B. Phase 2: End User Survey
The second phase refers to end-user survey used to collect
the MOS rates for 56 video sequences created in the first phase
according to ITU-T recommendation G.1011. The total of 15
examinees has participated in the study. Collected
demographic data related to age, gender, and educational level
of the examinees describes the group that approached the
questioning:
60% of examinees fit into the category of age 15 to 24,
13.34% into the category age 25 to 34, 13.34% into the
category age 35 to 44, and 13.33% into the category
age 44 to 55;
33% of examinees were male and 67% of them were
female;
20% of examinees reported as having a high-school
diploma, 60% reported as being students, 7% as having
faculty degree, and 13% of them as having a Master of
Science degree.
All participants were given a task of watching 56 test video
sequences by using Samsung Galaxy J5 smartphone after which
they were asked to express their opinion regarding their QoE
while using this video streaming service. The subjective
evaluation of test video sequences was performed by using the
electronic evaluation questionnaire, which contained the part
that was completed at the beginning of the experiment and it
included the questions that covered the information related to
the examinee’s personal data, previous experience with video
streaming service usage, and the part that deals with the
examinee’s rating of the statement related to overall QoE when
using video streaming service. The latter statement was a
simple MOS scale used as the de facto standard in QoE studies
and specified in ITU-T Recommendation P.800.1. It must be
noted that subjective QoE evaluation are to-date most
commonly reported in terms of a single MOS value [30].
The experiment procedure lasted about 30 minutes and has
included the following three steps [30]: (1) introduction and
clarification of the experiment tasks that need to be performed
by the examinee (5 minutes), (2) examinee training (5
minutes), and (3) testing and rating of video sequences (20
minutes). Video sequences have been displayed in landscape
mode and in the same order to all examinees. All examinees
were asked not to think about their feelings during evaluation,
but to be intuitive.
IV. RESULTS AND DISCUSSION
In order to address the set of hypotheses (H1-H7), stating
that difference in QoE caused by the impact of individual
parameters (i.e., routing protocol type, number of nodes, and
speed of nodes) and their interaction in MANETs are not
statistically significant, we decided to use the three-way
ANOVA [31]. After it has been determined that the considered
data satisfies the presumptions of ANOVA, i.e., normally
distributed variables, independent observations, and
homogeneity, we have performed a three-way ANOVA
analysis for MOS as a dependent variable. Independent
variables are protocol type, number of nodes, and speed of
nodes. The analysis has been performed by using the trial
version of IBM SPSS (Statistical Package for the Social
Sciences) Statistics 20 software [32].
The results of three-way ANOVA (Fig. 1) for MOS show
that there exists statistically significant interaction of protocol
type, number of nodes, and speed of nodes with large effect
size in terms of practical significance (F(1,14)=16.101,
p<0.001, =0.24). Practical significance is measured with
effect size , i.e., =0.032 - small, = 0.060 - medium,
and = 0.14 - large.
The existence of this statistically significant interaction
conditioned the conduction of two-way ANOVA analysis for
all parameters. After conducting the two-way ANOVAs the
following results are obtained.
There exists statistically significant interaction with large
effect size between number of nodes and their speed for AODV
protocol (F(1,14)= 31.966, p<0.001, =0.579), but as well
for DSDV protocol (F(1,14)= 13.238, p<0.001, =0.368).
Statistically significant interactions led to further consideration
of individual impacts of these factors on QoE.
Fig. 1. Graphical representation of the impact of three-way interaction of
protocol, number of nodes, and speed of nodes on MOS.
978-1-5386-4907-7/18/$31.00 ©2018 IEEE
By using post-hoc analysis, we have obtained the following
results. In first case, for node number value of 20 (F(1,6)=
44.636, p<0.001, =0.732) higher node speed leads to lower
MOS, i.e., QoE. For the remaining values, i.e., 30, 35, and 40
higher node speed leads to higher MOS (F(1,6)= 65.86,
p<0.001, =0.801; F(1,5)= 28.969, p<0.001, =0.636);
F(1,4)= 25.372, p<0.001, =0.592). In case of DSDV
protocol, for each node number value, i.e., 20, 30, 35, and 40
higher node speeds leads to higher MOS (F(1,6)= 7.155,
p<0.001, =0.305; F(1,6)= 14.727, p<0.001, =0.474;
F(1,6)= 2.865, p<0.001, =0.149; F(1,4)= 51.63, p<0.001,
=0.747).
Also, there exists statistically significant interactions with
large effect size between protocol and node speed for each
number of nodes (F(1,6)= 18.773, p<0.001, =0.365; F(1,6)=
38.545, p<0.001, =0.541; F(1,5)= 24.113, p<0.001,
=0.400; F(1,3)= 24.637, p<0.001, =0.346). As well as in
previous case, the statistically significant interactions led to
further consideration of individual impacts of these factors on
QoE. By using post-hoc analysis, we have obtained the
following results. In case when we had 20 nodes, regardless of
the node speed DSDV MOS rates (F(1,5)= 28.969, p<0.001,
=0.636) were higher than the ones for AODV (F(1,6)=
7.155, p<0.001, =0.305). In all other cases, MOS rates when
AODV was used were higher (F(1,6)= 65.86, p<0.001,
=0.801; F(1,5)= 28.969, p<0.001, =0.636; F(1,4)=
25.372, p<0.001, =0.592) than the ones in DSDV protocol
case regardless of node speed (F(1,6)= 6.432, p<0.001,
=0.474; F(1,6)= 2.865, p<0.001, =0.149; F(1,4)= 51.63,
p<0.001, =0.747).
Finally, going further with the investigation, we have found
that there exists also statistically significant interactions with
large effect size between protocol and number of nodes for each
node speed, except for the first one which has been
characterized as not significant (F(1,2)= 0.693, p=0.503,
=0.016; F(1,2)= 30.861, p<0.001, =0.269; F(1,3)= 8.215,
p<0.001, =0.182; F(1,3)= 33.185, p<0.001, =0.471;
F(1,2)= 6.295, p=0.003, =0.114; F(1,3)= 43.928, p<0.001,
=0.541; F(1,3)= 32.584, p<0.001, =0.466). Considering
the absence of significant interaction in case of the lowest node
speed, lead to analysis of main effects of individual variables
which show that their impact on MOS was statistically
significant with large effect size (F(1,1)= 19.019, p<0.001,
=0.185; F(1,2)= 19.712, p<0.001, =0.319), respectively
for protocol and number of nodes. As well as in the previous
case, the statistically significant interactions led to further
consideration of individual impacts of these factors on MOS
rates. By using post-hoc analysis, we have obtained the
following results. Regardless of node speed and protocol used,
number of nodes has shown to be influential with statistically
high significance and large effect (F(1,1)= 40.139, p<0.001,
=0.537; F(1,1)= 41.039, p<0.001, =0.687; F(1,3)=
10.055, p<0.001, =0.354; F(1,3)= 61.333, p<0.001,
=0.767; F(1,3)= 36.680, p<0.001, =0.663; F(1,3)=
41.963, p<0.001, =0.692; F(1,3)= 66.103, p<0.001,
=0.780; F(1,2)= 17.717, p<0.001, =0.458; F(1,3)=
82.018, p<0.001, =0.815; F(1,3)= 6.445, p=0.001,
=0.257; F(1,3)= 8.543, p<0.001, =0.314; F(1,3)= 35.803,
p<0.001, =0.657).
Based on the presented results, i.e., statistically significant
impacts and interactions of MANET-specific parameters on
QoE in terms of MOS rates when using video streaming
service, has shown that our hypotheses (H1-H7) are rejected
meaning that our expectations related to the existence of those
influences have been met. These results are valuable given the
fact that challenging deployment of real-time services over
MANETs is in fast lane and high QoE for them needs to be
achieved as well. Therefore, a comprehensive understanding of
MANET-specific parameters, impacts, as well as their
interactions, is a clear prerequisite for successful management
and optimization of end-user experience.
V. CONCLUSION AND FUTURE WORK
The influence of MANET-specific parameters on user
experience and its quality for video streaming has been
investigated not nearly as much as their impact on network
related SIFs, such as delay, jitter, loss, throughput, etc. Here
lies the motivation for this paper which tries to provide a
deeper understanding of the influence of MANET-specific
parameters on QoE. The aim has been to experimentally
investigate the existence of impact of MANET-specific
parameters on user's overall QoE in case of video streaming
service.
In order to accomplish this, a brief survey of the state-of-
the-art literature in the field of MANET-specific parameters
and their impact on network-related SIFs (i.e., delay, jitter,
loss, throughput, etc.) and QoE has been prepared. Further on,
experimental study has been conducted to obtain data for
analyzing the influence of selected MANET-specific
parameters, i.e. routing protocol type, number of nodes, and
speed of nodes, on user’s QoE when using video streaming
service. The statistical analysis of data collected within the
experiment implies that there exist strong and statistically
significant impact of individual parameters (i.e., routing
protocol type, number of nodes, and speed of nodes) and their
interaction in MANETs on QoE when using video streaming
service.
However, the conducted experimental study has certain
limitations that may be overcome in the future work. Giving
the fact that video streaming over MANETs is probably
influenced by additional parameters except those three
investigated in this paper, a broader range of various
parameters and their interactions should be included in future
research studies. Moreover, larger number of examinees
should be included in the further studies in order to draw non-
misleading conclusions. Since examinees involved in this
study were instructed on what task to perform, the field studies
should be conducted in the future, where examinees will not
978-1-5386-4907-7/18/$31.00 ©2018 IEEE
be guided in terms of performed task. In addition, the subject
of future work should be proposal of predictive
multidimensional model for video streaming over MANETs
which may be helpful in a practical sense.
REFERENCES
[1] E. Cero, J. Baraković Husić, and S. Baraković, “IoT’s Tiny Steps
towards 5G: Telco’s Perspective,” Symmetry, 2017, vol. 9, no. 10, 2017.
[2] J. M. Eseves da Silva, “Simulation of Algorithms for Mobile Ad-Hoc
Networks,” M.S. thesis, Departamento de Ciencia de Computadores,
Faculdade de Ciencias da Universidade do Porto, Porto, Portugal, 2015.
[3] G. Jisha, P. Samuel, and V. Paul, “Role of Gateways in MANET
Integration Scenarios,” Indian Journal of Science and Technology, vol.
9, no. 3, pp. 1-9, 2016.
[4] M. Lindeberg, S. Kristianse, T. Plagemann, and V. Goebel, “Challenges
and techniques for video streaming over mobile ad hoc networks,”
Mutimedia Systems, vol. 17, no. 1, pp. 55-82, 2011.
[5] S. Baraković and L. Skorin-Kapov, “Survey of research on Quality of
Experience modelling for web browsing,” Quality and User Experience,
vol. 2, no. 6, 2017.
[6] P. Le Callet, S. Möller, and A. Perkis, “Qualinet White Paper on
Definitions of Quality of Experience,” European Network on Quality of
Experience in Multimedia Systems and Services (COST Action IC 1003)
Version 1.2, Lausanne, Switzerland, 2013.
[7] S. Baraković, J. Baraković, and H. Bajrić, “QoE Dimensions and QoE
Measurement of NGN Services,” in Proc. of the 18th
Telecommunication Forum 2010 (TELFOR 2010), Belgrade, Serbia,
2010.
[8] R. Stankiewicz and A. Jajszczyk, “A Survey of QoE Assurance in
Converged Networks,” Computer Networks: The International Journal
of Computer and Telecommunications Networking, vol. 55, no. 7, pp.
1459-1473, 2011.
[9] L. Skorin-Kapov and M. Varela, “A Multi-Dimensional View of QoE:
the ARCU Model,” in Proc. of the 35th Jubilee International Convention
of Information and Communication Technology, Electronics and
Microelectronics (MIPRO 2012), Opatija, Croatia, 2012.
[10] D. D. Perkins, H. D. Hughes, and C. B. Owen,“Factors Affecting the
performance of Ad Hoc Netowrks,” in Proc. of the International
Conference on Communications, New York, USA, 2002.
[11] A. B. Malany, V. R. Sarma Dhulipala, R. M. Chandrasekaran,
“Throughput and Delay Comparison of MANET Routing Protocols,”
International Journal of Open Problems in Computer Science and
Mathematics, vol. 2, no. 3, pp. 462-468, 2009.
[12] S. Hakak, S. A. Latif, F. Anwar, and M. K. Alam,“Impact of Key
Factors on Average Jitter in MANET,” in Proc. of the 1st International
Conference on Systems Informatics, Modelling and Simulation,
Sheffield, United Kingdom, 2014.
[13] S. Hakak, F. Anwar, S. A. Latif, G. Gilkar, and M. K. Alam,“Impact of
packet size and node mobility pause time on average end to end delay
and jitter in MANETS,” Proc. of the International Conference on
Computer & Communication Engineering 2014, Chennai, India, 2014.
[14] S. S. Srivastava, B. Ram, A. Kumar,“Minimization of the Packet Losses
in MANETS base od both Static and Dynamic Routing Protocols,” in
Proc. of the International Symposium on Devices MEMS, Sikkim, India,
2011.
[15] M. K. Gulati and K. Kumar, “Performance Comparasion of Mobile Ad
Hoc Network Routing Protocols,” International Journal of Computer
Networks and Communications, vol.6, no.2, pp.127-142, 2014.
[16] M. Amnai, Y. Fakhri, and J. Abouchabaka, “Impact of Mobility on
Delay-Throughput Performance in Multi-Service Mobile Ad-Hoc
Networks,” International Journal of Communications, Network and
System Sciences, vol. 4, pp.395-402, 2011.
[17] M. Kumar and A. K. Jain, “Effect of Delay And Jitter on Wireless Ad-
Hoc Networks (IEEE 802.11) with Various Routing Protocols,”
International Journal of Engineering Research and Technology, vol. 2,
no. 6, pp.1392-1398, 2013.
[18] S. Baraković, S. Kasapović, and J. Baraković, “Comaprison of MANET
Routing Protocols in Different Traffic and Mobility Models,” Telfor
Journal, vol. 2, no. 1, pp. 8-12, 2010.
[19] R. K. Jha and P. Kharga, “A Comparative Performance Analysis of
Routing Protocols in MANET using NS3 Simulator,” International
Journal of Computer Network and Information Security, vol. 4, pp. 62-
68, 2015.
[20] S. Shrivastava, L. Shrivastava, and S. S. Bhadauria, “Performance
Analysis of Wireless Mobile ad Hoc Network with Varying Trasmission
Power,” International Journal of Sensor and Its Applications for Control
Systems, vol.3, no.1, pp. 1-6, 2015.
[21] T. R. Sheltami , E. M. Shakshuki, and H. T. Mouftah,“Video streaming
application over WEAC protocol in MANET,” Journal of Computer and
System Sciences, vol. 77, no. 4, pp. 705-719, 2011.
[22] P. T. A. Quang, K. Piamrat, K. D. Singh, and C. Viho,“Video Streaming
over Ad-hoc Network: a QoE- based Optimal Routing Solution,“ IEEE
Transaction on Vehucular Technology, vol. 66, pp. 1533-1546, 2016.
[23] M. S. Islam, A. Riaz, and M. Tarique, “Performance analysis of routing
protocols of mobile ad hoc networks for VoIP applications,” Cyber
Journals: Multidisciplinary Journlas in Science and Technology, Journal
of Selected Areas in Telecommunication, pp.26-33, 2012.
[24] P. Sondi, D. Gantsou, and S. Lecomte, “Performance Evaluation of
Multimedia Applications over an OLSR-based Ad Hoc Network using
OPNET,” in Proc. of 12th International Conference on Computer
Modelling and Simulation, Cambridge, United Kingdom, 2010.
[25] S. Mahajan and V. Chopra, “Performance Evaluation of MANET
Routing Protocol with Scalability using QoS Metrics of VoIP
Applications,” International Journal of Advanced Research in Computer
Science and Software Engineering, vol. 3, pp. 150-156, 2013.
[26] V. Goyal, S. Rani, and P. Singh, “Performance Investigation of Routing
Protocols for Database and Voice Data in MANETs,” International
Journal of Emerging Trends and Technology in Computer Science, vol.
2, pp. 326-332, 2013.
[27] S. Rubel. (March, 2017). Videezy. [Online]. Available at:
https://www.videezy.com/
[28] FFmpeg developer team. (July, 2011). ffmpeg. [Online]. Available at:
http://www.websabi.it/download/ffmpeg-doc.pdf
[29] NS-3 development team. (February, 2017) NS3. [Online]. Available at:
https://www.nsnam.org/docs/manual/ns-3-manual.pdf
[30] S. Baraković and L. Skorin-Kapov, “Multidimensional Modelling of
Quality of Experience for Mobile Web Browsing,” Computers in Human
Behaviour, vol. 50, pp. 314-332, 2015.
[31] T. C. Urdan, Statistics in plain English, 3rd ed., New York, NY:Taylor
and Francis Group, LLC, 2010.
[32] IBM development team. IBM SPSS Software. [Online]. Available at:
http://www-01.ibm.com/software/analytics/spss/
