See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/344465515

HTTP and FTP Application Algorithms to Identify the Optimal Network

Architecture and WLAN Protocol

Article in Solid State Technology · October 2020

CITATIONS READS

0 568

3 authors:

Ali mohd ali Mahmoud Dhimish

University of Huddersfield The University of York
11 PUBLICATIONS 23 CITATIONS 101 PUBLICATIONS 2,348 CITATIONS

SEE PROFILE SEE PROFILE

Peter J. Mather
University of Huddersfield
126 PUBLICATIONS 1,139 CITATIONS

SEE PROFILE

All content following this page was uploaded by Ali mohd ali on 24 December 2021.

The user has requested enhancement of the downloaded file.

HTTP and FTP Application Algorithms to Identify the Optimal Network Architecture and
WLAN Protocol

Ali Mohd Ali 1*, Mahmoud Dhimish 1, Peter Mather 1

1
Department of Engineering and Technology, University of Huddersfield, Queensgate HD1 3DH, Huddersfield, United
Kingdom
*
[email protected]

Abstract: This paper evaluates the performance of the HTTP Protocol and the FTP Protocol using OPNET simulation
(Riverbed Modeler 17.5). In this paper, an algorithm scheme is intended to examine HTTP and FTP services of different IEEE
802.11 technologies in order to define the optimal network architecture between the Basic Service Set (BSS), the Extended
Service Set (ESS) and the Independent Basic Service Set (IBSS). In addition, the proposed algorithm considers multi-criteria
network access selection, such as spatial distribution and number of nodes, to enable the provision of the best overall network
output and high-quality services. The Quality of Service (QoS) metrics used were delay, throughput and loss of packets. This
also meets the approval threshold values for HTTP and FTP quality metrics. This algorithm was applied to various room sizes
ranging from 2x3 m to 10x14 m and the number of nodes ranged from one to forty.

Keywords: HTTP, FTP, QoS, Performance Analysis; IEEE technologies.

1. Introduction
HTTP and FTP are best-effort services that are considered traffic is scheduled into the delay-conservative queue,
to be Internet-based services, where data packets are achieved a shorter delay.
forwarded to the network layer without any guarantee or Furthermore, a QoS algorithm was proposed by Chen et al.
preference for reliability or timely delivery, which is currently [5] to reduce the average delay time and jitter for VoIP
a major challenge for the communications industry as it has application and the packet loss ratio for high-definition video.
seen continued exponential growth. Best effort protocols are A. Mohd Ali et al. [6] aimed to construct different scenarios to
used to request/response pattern, and also work in TCP evaluate the characteristics of QoS and to examine the impact
(Transmission Control Protocol) protocols which is an of QoS enhancements. The evaluation, carried out using the
equitably good service for all file transfer applications. File OPNET simulator, would involve the various parameters of
transfer, either explicit (FTP) or implicit such as web page the WirelessLAN802.11e in order to see whether this
download (HTTP) or E-mail (SMTP), constitutes the improvement in the distribution of channel access improves
enormous majority of Internet traffic. A set of traffic the efficiency of the Wireless LAN 802.11 standard. The
measurements reported in Claffy et al. [1] suggests that 95% article by Wei et al. [7] studied the performance of HTTP and
of IP traffic is TCP, of which HTTP is 70%, FTP 5% and FTP protocols under the same network environment for five
SMTP 5%. With the rapid movement of business infrastructure clients. The study was conducted using two metric parameters
and home users towards Wireless LAN (WLAN), it is vital to average queuing delay and TCP delay and showed that the
implement their traffic over WLAN. WLAN has become performance of the HTTP protocol is better than the FTP
popular these days because it is easy and simple to deploy [2]. protocol. Seytnazarov and Kim [8] showed that in order for
Internet-based services such as web, email and file real-time services to work adequately, the QoS parameters and
transfers affect the usage of WLANs in addition to voice over characteristics performance have to be fulfilled and
wireless networks. Exchanging traditional data such as news, demonstrated that on the 802.11n network configured over 20
text applications, and file transfer have been successful using nodes the total throughput decreased. A. Mohd Ali et al. [9]
the Internet architecture. However, providing these services is proposes to use an algorithmic and mathematical scheme to
placing high demands on Internet architecture in both the allow the user/client to assess the optimum WLAN technology
number of connected hosts and the capacity of its data links . and the performance of the network architecture for a given
This, in turn, affects the quality of the service and is mix of internet applications such as HTTP, FTP and Email.
particularly noticeable when WLAN is used, resulting in poor For each application, the QoS metrics were adopted to develop
network performance [3]. In WLANs where multi- a computational algorithm model to provide accurate
applications have been deployed, a number of factors that numerical results.
affect the network performance should be addressed and Various efforts have been developed to evaluate the
evaluated such as the wireless network architectures (BSS, applications for QoS metric parameters that are configured
ESS and IBSS) and IEEE MAC layer technologies. over IEEE technologies [10-14]. QoS parameters such as an
However, providing precise QoS is considered as an issue end to end delay and throughput were observed by Sharma et
for best-effort services in the existence of real-time multimedia al. [10] across two IEEE technologies 802.11, 11g and
applications and has been the object of wide research [4-9]. demonstrated that the IEEE 802.11a technology performed
Firoiu [4] produced a novel architecture realized with a better across BSS network architecture. Mehmood and Alturki
combination of scheduling and queue management [11] introduced an architecture that analysed an IBSS network
mechanisms that classify WEB/TCP traffic as the drop- for a mix of HTTP, voice and video applications over 802.11g
conservative queue achieving a lower loss, and VoIP/UDP technology to scale and provisions QoS. This architecture
scales well with an increase in the network size and
outperforms well-known routing protocols. In addition, The subsequent sections of this article are organized as
Circiumarescu et al. [12] produced a comparative performance follows. Section 2 introduces the fundamentals and principles
analysis to ascertain which protocol among RIP, OSPF, of IEEE physical layer technologies. Section 3 presents the
EIGRP and IGRP best suits the network. This study was details of the proposed algorithm along with mathematical
conducted using QoS metric parameters such as packet delay calculations. In section 4 the results are analysed and evaluated
variation, packet end-to-end delay and video traffic analysis to in detail, while section 5 and 6 present a comparative study and
evaluate VC, E-mail, FTP and HTTP services using OPNET the conclusion.
and showed that the protocol that best suits VC is EIGRP.
Pérez et al. [13] introduced a scenario to evaluate IEEE 2. Preliminaries
802.11e standard for a number of videos, voice and best effort 2.1 IEEE MAC layer technologies
nodes, varying from 5 to 45 nodes, and showed an increase in
average delay for these services. Lakrami et al. [14] proposed The Institute of Electrical and Electronics Engineers
a new algorithm over infrastructure wireless network to (IEEE) developed the 802.11 group as a technology for
enhance the IEEE 802.11e in order to improve the QoS for WLAN technology. IEEE 802.11a operates in the 5 GHz
voice and video services which gives better results for all frequency band and 802.11b operates in the frequency band
performance metrics. 2.4 GHz, IEEE 802.11b supports transmission speeds of up to
To the best of our knowledge, no previous work has 11 Mbps and IEEE 802.11a provides a transmission speed of
evaluated the HTTP and FTP services QoS metrics of different 54 Mbps [16]. IEEE 802.11g supports transmission speeds of
IEEE 802.11 technologies in order to identify the optimum up to 54 Mbps by applying Orthogonal Frequency Division
technology standard across infrastructure and independent Multiplexing (OFDM) in the 2.4 GHz band. IEEE 802.11n
network architectures, which will be introduced in this article. uses Multiple Input Multiple Output Orthogonal Frequency
The implementation of QoS parameters such as delay, jitter Division Multiplexing (MIMO-OFDM) techniques to achieve
and packet loss over these best-effort networks is also transmission speeds of up to 300 Mbps. In case of using a
considered as an enormous challenge. At the same time, the channel bandwidth up to 40 MHz, IEEE 802.11n can provide
existence of different IEEE 802.11 technologies requires a transmission speeds of up to 600 Mbps [17]. IEEE 802.11
logical analysis to decide which technology should be used and standard does not support time-sensitive voice applications but
put into practice. Furthermore, the availability of IBSS, BSS only best-effort services. After several refinements and with
and ESS have increased the difficulty of deciding which the increasing call for real-time multimedia applications, a new
network architecture is best to use, regarding the assigned amendment named IEEE 802.11e was designed [18]. Table 1
wireless network resources, to provide optimum network shows the main differences between the IEEE 802.11
quality. Moreover, as demonstrated in [15] the optimum standards.
performance of IEEE technologies deployed in real-time Table 1 Summary of IEEE 802.11 standards
industrial communication systems not always guaranteed to Standard 802.11 802.11a 802.11b 802.11g 802.11n
recent technologies (802.11n) over the older one (802.11g), for MAC DCF DCF DCF DCF EDCA
this exact reason our work provides an analysing study that protocol
suggests to the user the optimum technology/technologies and Data Rate 1, 2 Up to 1, 2, 5.5, Up to Up to
network architecture without wasting resources nor getting in (Mbps) 54 11 54 600
the issues of randomly choosing specific technologies then
redesigning the whole configuration. Modulation FHSS, OFDM DSSS ERP- MIMO-
DSSS OFDM OFDM
On the other hand, the "technology", that is the PHY layer Frequency 2.4 5 2.4 2.4 2.4 & 5
is dependent on the hardware used and more capable hardware Band
can automatically downgrade the PHY layer for a lower (GHz)
throughput and better robustness when the signal-to-
Channel 20 20 20 20 20, or
interference-plus-noise ratio (SINR) is too low. This means Width 40
that devices can automatically switch from 11g to 11b for (MHz)
instance if the conditions are bad. But still can’t upgrade itself
automatically from 11g to 11e for instance. As an example, if Number of 1 1 1 1 1, 2, 3,
Spatial or 4
the devices are valid for a certain bandwidth and it's required Streams
to upload 8 Mbps video, at this point the device can’t upgrade
itself to match the required bandwidth, so either it is needed to 2.2 IEEE networks infrastructures
upgrade it physically or downgrade the video quality to IEEE 802.11 defines two basic modes of
optimize the available bandwidth. However, this is exactly communication between WLAN nodes: Infrastructure and
where this study is beneficial and provide its main contribution, Independent which are known as Ad Hoc Networks [19].
that it maintains the resources (cost-efficient) and provide Infrastructure BSS is a group of stations that connect to the
network optimization, it is not only considered the usage by its same wireless medium and are controlled by a centralized
own. Furthermore, the availability of IBSS, BSS, and ESS coordination function or access point (AP). All stations can
have increased the difficulty of deciding which network communicate directly with all other stations in a fixed range of
architecture is best to use, regarding the assigned wireless the base station. The IEEE 802.11 infrastructure networks use
network resources, to provide optimum network quality. APs. AP supports wave extension by providing the integration
This article looks into the possibilities of having any effects points necessary for network connectivity between multiple
on network performance when using a different number of BSSs, thus forming an Extended Service Set (ESS). In addition,
nodes and IEEE physical layer technologies implemented the IBSS or Ad-hoc network is a specified group of nodes in a
across various spatial distributions.
single BSS for the purpose of internet working without the aid successful technology companies use the model to develop
of a centralized coordination function [20] (i.e. access point). their research and development processes.
Using OPNET simulation, we have considered two main
2.3 QoS performance metrics and Importance coefficient sources’ inputs for this algorithm: user configurations and
for HTTP and FTP applications technology specifications (standards). User configurations
define the number of nodes that are needed in the network and
Performance metrics are defined in terms of QoS metric spatial distribution. Technology specifications (standards)
parameters for multi-service applications. For each application, define the physical layer technologies and network
a satisfaction criterion (acceptable threshold) for each QoS architectures.
metric parameter is identified [21, 22] as shown in Table 2, The top part of Fig. 1 defines these factors. Network
which represents the key QoS requirements and architectures specify how different wireless components
recommendations for each application (bearer traffic). connect together in either of two modes: the presence of access
Table 2 QoS metric parameters importance for HTTP and points (BSS and ESS) mode or the absence of access points
FTP applications (IBSS) mode, number of nodes needed in this network which
Application Importance Delay/ Throughput Racket
breaks down to four groups (0-5, 6-10, 11-20 and 21-40),
& Threshold Respon (kbps) Loss spatial distribution which specifies the topology in which these
se time Rate nodes will be distributed − in a circular (oval) way, uniform
(sec) (%) (grid) way, or randomly scattered way. IEEE MAC

Importance M L L
HTTP
Threshold 1 30 10

Importance L M H
FTP
Threshold 1 45 5

Where: H=High, M=Medium and L=Low

HTTP and FTP applications quality is directly affected by

the following QoS metric measurements:
• Packet End-to-End delay (sec): the time taken by
data/voice to travel from node A to node B on the
network.
• Page response time (sec): the time required to retrieve
the entire page with all the contained inline objects.
• Throughput (bit/sec): the total rate at which packets are
transferred from the source to the destination at a
prescribed time period.
• Traffic Sent (packet/sec) and Traffic Received
(packet/sec): used to calculate packet loss rate, which is
the percentage of packets that get lost along the
communication path after the packet is transmitted by
the sender into the network.
It is worth noting that an Important Coefficient is
assigned to each Application parameters (ICA) in terms of its
impact on the call quality of the service. Table 2 shows the
QoS qualitative importance of each QoS parameter and their
related threshold values for each application. In order to be
able to account for these qualitative factors in a simulation they
have to be translated into numbers (H=1, M=0.5, L=0.1, and
VL=0).

3. Proposed algorithm: Protocol and network architecture

selection
3.1 Building projects (Simulation environment)
In this paper, an OPNET simulation platform [23] is used
to build and analyse all applications scenarios. OPNET is a
discrete event system simulator that simulates the system
behaviour by modelling each event happening in the system
and processing it by a user-defined process. OPNET Modeler
allows you to study communications networks, equipment, Fig. 1. Flowchart of the proposed algorithm
applications and protocols with ease and scalability. The most
Technologies defines the physical layer technologies that will Table 4 FTP Simulated traffic parameters
be used to build many different scenarios.
Parameters Values
All network architectures (BSS, ESS, IBSS) have been
configured and implemented across all three spatial Command Mix 50%
distributions (circular, uniform, random) for the four groups of (Get/Total)
nodes. Figs. 2(a), (b) and (c) show some of these implemented Inter-Request Time Exponential (360)
scenarios. (sec)
The performances of different scenarios for each File Size (bytes) 50000
application have been investigated via an OPNET simulator. Types of service (TOS) Best Effort
The IEEE MAC layer technologies used were 802.11 (FHSS),
802.11a (OFDM), 802.11b (DSSS), 802.11g (OFDM) and
802.11e (QoS). The protocols used and the multi-service
A considered number of nodes up to 40 is in line with the
applications’ settings for the simulation are listed in Tables 3
literature [7, 8, 11 and 13]. On the other hand, it was found that
and 4.
all observed results obtained using these four groups of nodes
Table 3 HTTP Simulated traffic parameters are suitable to maintain of the network performance’s quality
that is a larger number of nodes in the network implies that
Parameters Values
relatively small amount of traffic will cause performance
HTTP Specification HTTP 1.1 degradation in the network due to fixed network bandwidth
Page Interval Time Exponential (60) capacity.
(sec)
Types of service (TOS) Best Effort

a b

Fig. 2. Design of the three Network Architectures across three Spatial Distributions for HTTP and FTP
(a) Basic Service Set (BSS), (b) Extended Service Set (ESS), (c) Independent Basic Service Set (IBSS)
3.2 System model’s calculation distribution F(n) [24] is going to be produced for all
The system calculations and the mathematical model applications QoS metric parameters from the OPNET Modeler
are shown in phase II at the bottom part of Fig. 1. The inputs simulation, then analysed against PTV as follows:
for the algorithm’s mathematical calculations are QoS 1. If ptv ∈ F(n): it means that the PTV has a specific
Threshold values for each application and Cumulative value on its CDF distribution equal to QPM for this
Distribution Function (CDF) distribution. Applications QoS metric parameter. QPM is weighted by ICA to
Threshold values (satisfaction criterion) are taken from produce QFM. Then the aggregation of all QFMs
literature as shown in Table 2 [21, 22]. CDF distribution is yields AFM which is used to classify IEEE
produced for these QoS metric parameters from OPNET after technologies.
running the simulation scenarios.
2. If ptv > F(n): it means that the QPM value equals 1
Mathematical calculations will be done to determine how a
and QFM has arisen.
particular scenario has satisfied certain performance metrics
for each application. The following steps are used to explain 3. If ptv < F(n): it means that the QPM value equals 0
the calculations of this algorithm and to analyse the results for and QFM will be initialized.
each of the above projects. The value generated for the applications QoS metric
• QoS Performance Metric (QPM): as Fig. 3 illustrates, parameters (jitter, delay, throughput and packet loss) will
the value that is produced by applying the application contribute to filling in Table 5 which leads to a rank order of
QoS metric Parameter Threshold Value (PTV) for IEEE technologies for each network architecture.
each QoS performance criterion n once is represented All applications QoS metric parameters will be calculated
in CDF distribution F(n), which is given by (1). as explained in the previous sections except for a packet loss
parameter. OPNET Modeler is designed to produce the result
𝑄𝑃𝑀𝑛 = 𝐹(𝑝𝑡𝑣) (1) of the packet loss parameter as a Boolean value (0.0 or 1.0)
that corresponds to the acceptance or rejection of a packet,
1 respectively. However, this work requires a numerical value
for the packet loss.
CDF probability

0.8
A code has been programmed using MATLAB software
0.6 QPMj to develop a method to calculate the packet loss percentage for
Time response
0.4 threshold for FTP
each application. This method is linked directly with the
(PTVj) OPNET Modeler to produce a specific packet loss percentage
0.2 for each application. Application packet loss rate 𝜔𝑖 of a node
0 i is the ratio of dropped voice packet 𝑘i to total voice packets
0 0.5 1 1.5 2 𝜌𝑖 multiplied by 100%, as demonstrated by (4).
Response Time (sec) 𝜔𝑖 = ( 𝑘𝑖 ⁄𝜌𝑖 ) ∗ 100% (4)

This requires the traffic received/send rate values from

Fig. 3. QPM for Response time
OPNET Modeler to be integrated to produce the total number
• QoS Fitness Metric (QFM): the value that is produced by of packets received and sent. Then, the exact packet loss ratio
applying a weighting to the QPM (assigned by importance) is produced and should be presented as a CDF diagram to
for each QoS metric parameter (H=1, M=0.5, L=0.1, and enable identification of the values of QPM, QFM and AFM
VL=0) is expressed by (2). using the previously explained flowchart.
Identical calculation steps were applied for the other three
𝑄𝐹𝑀𝑛 = 𝑄𝑃𝑀𝑛 ∗ 𝐼𝐶𝐴 (2) groups of nodes (0-5, 11-20 and 21-40), to ascertain the best
performing IEEE technology/technologies and to produce all
• The final step will be calculating the Application Fitness
Metric (AFM) which is to aggregate all QFMs for n Table 5 IEEE technologies calculation and rank order list for one
application QoS metric parameters (delay, jitter, project
throughput and packet loss), for each IEEE 802.11 Application
technology j, as demonstrated by (3). Technolog
4 Technolo y
Delay/Res AFM
gy Throughp Packet Rank
𝐴𝐹𝑀𝑗 = ∑ 𝑄𝐹𝑀𝑛 (3) ponse
ut Loss order
Time
𝑛=1

• Based on AFMs of the IEEE 802.11 technologies, the

Technolog
802.11 QFMD QFMTH QFMPL AFM11
y1
rank order of these five technologies will be produced
AFM11 Technolog
for each of the three built network architectures. Hence, 802.11a QFMD QFMTH QFMPL
a y2
the best network architecture performance will be
AFM11 Technolog
identified for all groups of nodes as will be explained 802.11b QFMD QFMTH QFMPL
b y3
later in section 4.
AFM11 Technolog
802.11g QFMD QFMTH QFMPL
The flowchart presented in Fig. 1 illustrates these g y4
mathematical steps which produce the AFM value for each AFM11 Technolog
802.11e QFMD QFMTH QFMPL
IEEE MAC technology. As explained previously, CDF e y5
values of QPMs, QFMs, and AFMs for all QoS metric Both results’ flowcharts start by identifying the number of
parameters regarding each application in all network nodes that will be used to configure the required network and
architectures across the three spatial distributions. work for the environment composed of 1 to 40 nodes.

4. Results and performance evaluation 4.1 Results of HTTP

In this article, the output of the proposed algorithm Based on the user’s configuration and the number of
identifies the options available for a client (user) based on the nodes required to set up the designated network, both results’
tables of the results that have been produced for all scenarios algorithms classify four key groups of nodes, presented as
across three network architectures. By options, the best follows:
performing technologies across all three network architectures 1. The first, second and fourth categories, where 5 ≥ N >
(IBSS, BSS and ESS) is implied. The results are divided into 0, 10 ≥ N > 5 and 40 ≥ N > 20, respectively, in the
three main sections related to the both applications (HTTP and generic flowchart, as can be seen in Table 6, both
FTP). All simulated scenarios are applicable to the lab (room) architectures BSS and ESS provide the best
sizes from 2x3m to 10x14m. performance across all spatial distributions for all five
The format of the results is demonstrated based on the technologies. However, in the case of the IBSS
presence of an access point; therefore, the tables of the results flowchart for the first and second categories, the
are interpreted (translated) as: generic result and IBSS only, as 802.11 technology provides the best performance,
will be demonstrated later for each application. while all technologies performing well in the fourth
category, for all three spatial distributions as shown
• In case there is at least one access point in the network, then in Table 7.
the proposed algorithm in Fig. 1 and the result in Tables 6 2. The third category, where 20 ≥ N > 10, in the generic
and 8 will be applied. This case is applicable to both
flowchart, the BSS and ESS provide a number of
infrastructure architecture layers (ESS and BSS). All
options. For BSS architecture, the five technologies
scenarios are running in all five IEEE 802.11 technologies
perform well across all three spatial distributions.
and three spatial distributions: circular, uniform, and
IEEE 802.11, 11b, and 11g, are acknowledged as the
random.
preferable solutions for ESS. However, according to
• If the network is configured without any access points, then the IBSS flowchart, all IEEE technologies are
the proposed algorithm in Fig. 1 and the IBSS result’s performing well as can be seen in Table 7.
described Tables 7 and 9 will be used. All scenarios are
running in all five IEEE 802.11 technologies and three
spatial distributions: circular, uniform and random.

Table 6 BSS and ESS Generic Algorithm Results for HTTP

Application HTTP

User Configuration

Number of Nodes 5≥N>0 10 ≥ N > 5 20 ≥ N > 10 40 ≥ N > 20

Network Architecture BSS or ESS BSS or ESS BSS ESS BSS or ESS

System Specification

Spatial Distribution C U R C U R C U R C U R C U R

802.11 802.11 802.11 802.11 802.11

802.11a 802.11a 802.11a 802.11b 802.11a
802.11b 802.11b 802.11b 802.11g 802.11b
IEEE Technology
802.11g 802.11g 802.11g 802.11g
802.11e 802.11e 802.11e 802.11e
 Table 7 IBSS Algorithm Results for HTTP

Application HTTP

User Configuration

Number of Nodes 5≥N>0 10 ≥ N > 5 20 ≥ N > 10 40 ≥ N > 20

Network Architecture IBSS IBSS IBSS IBSS

System Specification

Spatial Distribution C U R C U R C U R C U R

802.11 802.11 802.11 802.11

802.11a 802.11a
802.11b 802.11b
IEEE Technology
802.11g 802.11g
802.11e 802.11e

4.2 Results of FTP

1. The first category, where 5 ≥ N > 0, in the generic table, all technologies perform well across all spatial
flowchart, as shown in Table 8, BSS is the best distributions.
architecture network. The five technologies perform 3. The third category, where 20 ≥ N > 10, as can be
well across uniform and random distributions. seen in Table 8, both architectures BSS and ESS
However, according to the IBSS flowchart, IEEE provide the best performance across all spatial
802.11b technology is considered the preferable distributions for all five technologies. While, for the
solution as shown in Table 9. fourth category, where 40 ≥ N > 20, the best network
2. The second category, where 10 ≥ N > 5, the BSS and architecture for this large network is ESS across all
ESS provide a number of options. BSS and ESS spatial distribution. However, according to the IBSS
provide optimum performance if the network is flowchart, all technologies perform well across all
configured circularly and uniformly, respectively, for spatial distributions for both categories, third and
all five technologies. However, according to the IBSS fourth, as shown in Table 9.
Table 8 BSS and ESS Generic Algorithm Results for FTP

Application FTP

User Configuration

Number of Nodes 5≥N>0 10 ≥ N > 5 20 ≥ N > 10 40 ≥ N > 20

Network Architecture BSS BSS ESS BSS or ESS ESS

System Specification

Spatial Distribution U R C U C U R C U R

802.11 802.11 802.11 802.11

802.11a 802.11a 802.11a 802.11a
802.11b 802.11b 802.11b 802.11b
IEEE Technology
802.11g 802.11g 802.11g 802.11g
802.11e 802.11e 802.11e 802.11e
 Table 9 IBSS Algorithm Results for FTP

Application FTP

User Configuration

Number of Nodes 5≥N>0 10 ≥ N > 5 20 ≥ N > 10 40 ≥ N > 20

Network Architecture IBSS IBSS IBSS IBSS

System Specification

Spatial Distribution C U R C U R C U R C U R

802.11b 802.11 802.11 802.11

802.11a 802.11a 802.11a
802.11b 802.11b 802.11b
IEEE Technology
802.11g 802.11g 802.11g
802.11e 802.11e 802.11e

generation networking”. International Journal of

5. Conclusion Communication Systems, 2011, 24(9), 1150-1162.
[6] A. Mohd Ali, M. Sibley, I. Glover. “WLAN 802.11e
This work has developed a novel algorithm to assess evaluation performance using OPNET”. International
HTTP and FTP applications QoS metrics of different IEEE Journal of All Research Education & Scientific Methods
802.11 technologies in order to choose the optimum network (IJARESM), 2017, 5(1), 35-38.
architecture among BSS, ESS and IBSS. The rank order of [7] P. Wei, Z. Hong, M.Shi. “Performance analysis of HTTP
different IEEE 802.11 technologies has been produced across and FTP based on OPNET”. In 2016 IEEE/ACIS 15th
different spatial distributions. The results of FTP applications International Conference on Computer and Information
show that it is only preferable to use the ESS network with a Science (ICIS) (pp. 1-4). IEEE.
high number of workstations/nodes in a network; this is due to [8] S. Seytnazarov, Y. Kim, “QoS-Aware Adaptive A-MPDU
Aggregation Scheduler for Voice Traffic in Aggregation-
the high packet loss and delay that might appear in the network Enabled High Throughput WLANs”. in IEEE Transactions
owing to the increase in the number of workstations. on Mobile Computing, vol. 16, no. 10, pp. 2862-2875, 1 Oct.
Additionally, almost all IEEE technologies work for all spatial 2017, doi: 10.1109/TMC.2017.2672994.
distributions. Furthermore, IBSS can be worked efficiently [9] A. Mohd Ali, M. Dhimish, I. Glover. “WLAN Protocol and
with all technologies in almost all network sizes. On the other Network Architecture Identification for Service Mix
hand, ESS architecture has the same performance for all spatial Applications”. International Journal of Advance
distributions regardless of the network size for HTTP Computational Engineering and Networking (IJACEN),
application. Notwithstanding BSS performance works well 2020, 8(2), 24-30.
with all number of nodes in case of HTTP. Furthermore, the [10] V. Sharma, J. Malhotra, H. Singh. “Quality of Service (QoS)
evaluation of IEEE 802.11 WLAN using different PHY-
results of HTTP show that IBSS performs well with 802.11 Layer Standards”. Optik-International Journal for Light
technology in small network sizes and with all technologies in and Electron Optics, 2013, 124(4), 357-360.
bigger ones. [11] Mehmood, R., & Alturki, R. “A scalable multimedia QoS
architecture for ad hoc networks”. Multimedia Tools and
Applications, 2011, 54(3), 551-568.
REFERENCES [12] L. D. Circiumarescu, G. Predusca, N. Angelescu, D.
[1] K. Claffy, G. Miller, K. Thompson. The nature of the beast: Puchianu. "Comparative Analysis of Protocol RIP, OSPF,
recent traffic measurements from an Internet backbone. In RIGRP and IGRP for Service Video Conferencing, E-mail,
Inet98, 1998. FTP, HTTP." 20th International Conference on Control
[2] J. F. Kurose, K. W. Ross. Computer Networking: A Top- Systems and Computer Science. IEEE, 2015.
Down Approach Featuring the internet, Addison Wesley, [13] S. Pérez, H. Facchini, A. Dantiacq, G. Cangemi, J. Campos.
2012. “An evaluation of QoS for intensive video traffic over
[3] C. Timmerer, H. Hellwagner. ‘‘Interoperable adaptive 802.11 e WLANs”. In Electronics, Communications and
multimedia communication’’. IEEE multimedia, 2005, Computers (CONIELECOMP), 2015 International
12(1), 74-79. Conference on (pp. 8-15). IEEE.
[4] V. Firoiu, X. Zhang, Y. Guo. “Best effort differentiated [14] F. Lakrami, M. El-Kamili, N. Elkamoun. “Towards the
services: Trade-off service differentiation for elastic Enhancement of QoS in 802.11 e for Ad Hoc Networks”.
applications”. IEEE ICT. Vol. 88. 2001. In Advances in Ubiquitous Networking 2, 2017, (pp. 95-
[5] J. L. Chen, S. W. Liu, S. L. Wu, M. C. Chen. “Cross‐layer 107). Springer, Singapore.
and cognitive QoS management system for next ‐
 [15] F. Tramarin, S. Vitturi, M. Luvisotto, A.Zanella, ‘‘On the
use of IEEE 802.11 n for industrial communications’’.
IEEE Transactions on Industrial Informatics, 2016, 12(5),
1877-1886.
[16] IEEE Std. 802.11-2007, Part 11: Wireless LAN Medium
Access Control (MAC) and Physical Layer (PHY), June
2007.
[17] IEEE Std. 802.11n-2009, Part 11: Wireless LAN Medium
Access Control (MAC) and Physical Layer (PHY)
Specifications Amendment 5: Enhancements for Higher
Throughput, Oct. 2009.
[18] F. Babich, M. Comisso, A. Crismani, "Considerations on
the multiplexing and diversity tradeoff in IEEE 802.11
networks," in IET Communications, vol. 8, no. 9, pp. 1551-
1559, 12 June 2014, doi: 10.1049/iet-com.2013.0741.
[19] A. Ksentini, “Quality of service (QoS) in IEEE 802.11-
based Wireless Local Area Networks (WLAN),” Cergy-
Pontoise University, 2005.
[20] P. Swain, S. Chakraborty, S. Nandi, P. Bhaduri,
"Performance Modeling and Analysis of IEEE 802.11 IBSS
PSM in Different Traffic Conditions," in IEEE
Transactions on Mobile Computing, 14.8, pp. 1644-1658, 1
Aug. 2015, doi: 10.1109/TMC.2014.2362754.
[21] N. Ghiata, M. Marcu. “Measurement methods for QoS in
VoIP review”. In Ultra Modern Telecommunications and
Control Systems and Workshops (ICUMT), 2011 3rd
International Congress on 2011 Oct 5 (pp. 1-6). IEEE.
[22] C. Hattingh, Szigeti, T. End-to-End QoS Network Design:
Quality of Service in LANs, WANs, and VPNs, 2004.
[23] Riverbed. (2018). Retrieved from Riverbed Web site:
https://www.riverbed.com/gb/index.html.
[24] R. D. Yates, D. J. Goodman. Probability and stochastic
processes: a friendly introduction for electrical and
computer engineers. John Wiley & Sons. 2014.

View publication stats