Impact of Randomness on MAC Layer Schedulers over High Speed Wireless Campus Network in IEEE802.11e

International Journal of Computer Science and Software Engineering (IJCSSE), Volume 1, Issue 1, October 2014
ISSN (Online): 2409-4285 www.IJCSSE.org Page: 8-13
Impact of Randomness on MAC Layer Schedulers over High Speed
Wireless Campus Network in IEEE802.11e
M. K. Alam1, S. A. Latif2, M. Akter3 and M. Y. Arafat4
1, 2, 3, 4 Department of Electrical and Computer Engineering, International Islamic University Malaysia,
Kuala Lumpur, Malaysia
1shishir_lmu@yahoo.com, 2suhaimie@iium.edu.my, 3m.tethi@gmail.com, 4yeasir.ete.ruet09@gmail.com
ABSTRACT
IEEE802.11e standard provides an EDCA mechanism to ensure
the QoS using service differentiation function for IMM traffics
through the WCN. In an EDCA mechanism, IMM traffics have
been classified into four categories and each category has
specific priority. Many EDCA schedulers allow these traffics to
access the channel based on assigned priorities. The IMM high
priority traffic is high delay sensitive and required high
bandwidth to ensure QoS over the networks. Scheduling
mechanism is the effective way to satisfy IMM traffic but it
cannot give guarantee because of the randomness of some EDCA
schedulers or proposed networks. In order to provide the QoS,
scheduling mechanism is the challenging issue to transmit IMM
traffic through the channel and it is still need to be solved over
WCN. This work aims to examine the effect of the randomness
on different MAC layer schedulers' performance and evaluates
the performance of the proposed schedulers using different
performance metrics in WCN.
Keywords: Scheduling Algorithms, Seed Values, WCN,
Multimedia Applications, Back-off Algorithm.
1. INTRODUCTION
Real time Interactive Multimedia (IMM) that is standard
application from Wireless Local Area Networks (WLANs)
and includes VoIP, Mobile Video conference and so on
from past years have shown tremendous growth and
deployment in IEEE802.11 [1-3].The ease of evaluation
for high speed Wireless Campus Network (WCN), a
interconnection of multiple WLANs (in a single wireless
subnet) has been made much easier by the contribution and
advancement of IMM applications [4]. Due to the
heterogeneities and constraints like- the requirements of
severe Quality of Service (QoS), limited bandwidth and
the different standards etc, the transmission of IMM
applications over the WCN is a challenging issue. Among
IMM applications , some major services include voice,
video conferencing, animation, graphics and so on over the
WCN and for optimal performance ,these applications
require minimum data loss , high throughput , least delay
and many other such factors [5].One of the effective way
to enhance the performance of the IMM traffics through
the WCN in IEEE802.11 is a technique known as Frame
scheduling [6].Besides , there has been some
enhancements made in the IEEE802.11e standard with
respect to QoS for IMM applications over the networks
[7]. The service differentiation based on assigning
priorities is the effective way to guarantee QoS which
provides IEEE802.11e standard EDCA protocol. Figure 1
shows the access mechanism of the different categories of
IMM traffics in EDCA protocol.
Multimedia Applications
AC0 AC1 AC2 AC3
Background Best Effort Video Voice
Fig. 1. Access Mechanism of multimedia traffic in IEEE802.11e based
on priority
An EDCA protocol is a contention based Medium Access
Control (MAC) layer protocol which differentiates the
IMM traffics based on various Arbitration Intra-Frame
Space Numbers (AIFSNs) and assigns the Transmission
Opportunity (TXOP) limit for high priorities traffics to
occupy the channel without other traffics interruption.
Table I displays the default parameters of an EDCA
protocol. These parameters values ensure to occupy the
wireless channel for a specific duration of time for
transmission without interruption over the network. After

9
M. K. Alam et al.
that, an EDCA mechanism cannot guarantee to deliver
IMM traffics with the least delay through the network due
to the randomness of some MAC layer schedulers and
mechanism of the proposed network.
Table 1: EDCA Default Parameters Set
Fig. 1. ACs Fig. 2. AIFS
N
Fig. 3. TXOPli
mit
Fig. 4. AC3(Voice) Fig. 5. 2 Fig. 6. 3.008ms
ec
Fig. 7. AC(Video) Fig. 8. 2 Fig. 9. 6.016ms
ec
Fig. 10. AC(Best
Effort)
Fig. 11. 3 Fig. 12. 0
Fig. 13. AC(Backgrou
nd)
Fig. 14. 7 Fig. 15. 0
Many researchers have worked to reduce the
randomness from the network because it could be
degraded the network performance unluckily but it is still
unresolved problem over the networks. This work aims to
examine the effect of the randomness on different MAC
layer schedulers' performance and evaluates the
performance of the proposed schedulers using different
performance metrics in WCN.
The rest of this paper is organized as follows: section II
deals with the background studies of different randomness
characteristics schemes, the proposed scheme of this paper
introduces in section III, section IV employs the
simulation results and discussion and this paper is
concluded with the last part.
2. BACKGROUND STUDIES
In this section, we review some related to random number
or position selection schemes and MAC layer scheduling
schemes that are very commonly used in any wireless
mobile network. Moreover, we discuss about the
importance of seed values. By using these seed values we
can generate different random numbers or positions that
are very important to analysis the network performance.
2.1 Back-off Window (BW) mechanism
Providing the QoS the EDCA protocol employed the back-off
window mechanism but it cannot provide guarantee
strictly due the randomness of the back-off window
algorithm. The back-off procedure indicates the current
channel status is idle or busy by sitting the random back-off
timer whose value is selected uniformly in the
[0,CWmin] interval. If the current medium becomes busy
during the back-off process then the node pauses it back-off
timer and waits until the channel is idle again. When
the back-off time will be expired then the next
transmission will be attempted. If the collision occurs the
back-off window mechanism increases the contention
window size exponentially to reduce the other collision.
During the first transmission attempt the contention
window is set the minimum contention window size after
every unsuccessful transmission it increases exponentially
until the contention window maximum [8]. Table II
highlights the different minimum and maximum
contention window sizes based on various priority traffics
in EDCA MAC layer protocol.
Table 2: Contention Window (CW) Sizes Based on Different Traffic
Priorities
Access
categories
(ACs)
Contention Window (CW) sizes
CWmin CWmax
AC3 7 15
AC2 15 31
AC1 31 1023
AC0 31 1023
2.2 Random Waypoint (RW) mobility model
The great impact of mobility models is observed in any
wireless mobile communication networks. The mobility
models are considered for simulation of any wireless
mobile network systems in order to justify the user
mobility on different distinctiveness of these systems [9].
There is lots of mobility models exist in modern wireless
technology [10], [11]. These models are employed in
different wireless networks. For instance, the random
waypoint mobility model is commonly used in any
wireless mobile networks [12]. In this model, every mobile
node of the wireless network selects the destination
randomly for moving during simulation. Let assumes the
mobile nodes are placed in the wireless network. During
the simulation these nodes move from one place to another
place randomly using the random waypoint mobility
model. When these nodes move and reach one destination
point after that these remain for static pause time and then
start to move again according to the same mobility rule
[13].
2.3 Random Number Generators
The random number generators determine the outcome of
any randomness schemes in any gaming devices or
networks. Typically, the random number generator is used
in many electronics gaming devices. It generates the
random numbers based on seed value (i.e., the outcome of
the random generator). The seed value can be any numeric
numbers even it can be represent by numeric date and time.

10
M. K. Alam et al.
Any random number generator algorithms are used the
seed values as an input of the algorithms that produce
different random numbers [14].
2.4 Seed Value
The importance of seed value is quick growth to analysis
the performance of the wireless mobile network and any
gaming devices that use random number generators. In any
random number generator algorithms can only generate the
static random numbers. They cannot change those
numbers. Using different seed value can be change the
random numbers of any random algorithm [15]. The
wireless network is used some random number or
destination selection mechanisms. The performance
analysis of this network used to change the random
number or position in any random mechanism which is
done by using various seed values.
2.5 Strict Priority Scheduler (SPS)
The SPS schedule the different categories of traffic based
on assigned priority. Arriving traffics in the SPS buffer are
classified in highest to lowest priority discipline manner in
the output queue. It provides service differentiation facility
according to assigned priorities to satisfy the QoS through
the networks. When the different categories of traffics in
the input link first assign the priorities on these then
transmit them to the medium according to the priority
levels. The highest priority traffics transmit through the
medium before serving the lowest priority traffics but if
the lowest priority traffics transmission ongoing over the
channel and on that time if the highest priority traffics
arrive in the scheduler input buffer then the highest
priority traffics have to wait until completed the current
lowest priority traffic served [16]. The main limitation of
SPS the lowest priority traffics may receive little attention
during the transmission of the highest priority traffic
streams.
2.6 First in First out (FIFO)
FIFO scheduling discipline specifies the traffics for
transmission over the channel in the same order in which
they reached at the output queue. In particular, people are
waiting to maintain a queue to get the bus in the bus stop.
Who comes first in queue he will get serve first to enter
the bus. Moreover, like customers join in the back of the
single waiting line and serve from front of the waiting line
[17]. Many researchers work on FIFO algorithm and they
have proved that the FIFO scheduler leads to a feasible
real time constraints schedule which minimizes the energy
consumption and gives the optimal schedule considering
shortest path problem with the minimal voltage and speed
changes[16]. Although, FIFO provides the mentionable
facilities over the networks but it cannot differentiate the
traffic categories into different priority classes. Moreover,
it cannot identify the ill-behaved traffic flow from the
other flows. Thus, the FIFO increases the end-to-end delay,
jitter and packets loss over the network.
2.7 Random Early Detection Scheduler (RED)
The RED scheduling mechanism is well known for an
active queue management and it is also known as random
early drop and random early discard algorithm. This is
employed to control the network congestion. It plays a
vital role over the network by the way of not allow to
admit full queues for further processing. As a result, it
reduces the frame delay and loss. According to the
statistical probabilities, the RED monitors the average size
of queue and the number of dropped packets’ queues. It
maintains the threshold based queuing manner.
Statistically, the RED drops the packets from the dropped
packet flows until reaching its threshold value. Hence, it
maintains a good packets queue for situations where per-session
state tracking’s complexity is needed. Many
researchers have employed to control the congestion,
resource management and so on in [18]. However, the
higher maximum threshold value increases the delay and
the minimum maximum threshold value decreases the
throughput over the networks.
2.8 Weighted Random Early Detection (WRED)
scheduler
WRED scheduling mechanism is the update version of
RED scheduler. Both mechanism control the congestion
over the network but WRED can avoid congestion more
effectively and improve the gateway’s queue when it is too
full for its buffer. The RED scheduler avoids congestion
by dropping the packets with the dropping probability.
When the average queue size is less than minimum
threshold and maximum threshold then the packets
dropped. However, the RED scheduler router has no any
bias against the bursty traffics hence, avoids the global
synchronization which reduces the congestion window of
many connections at the same time. The RED is very easy
to configure but it will not justifiably for the processing
resource of gateway. As a result, The RED cannot ensure
to provide optimal QoS for real time sensitive applications
to the end users. The WRED develops the congestion
avoidance mechanism for QoS sensitive applications. It
selects the packets for dropping based on either Resource
Reservation Protocol (RSVP) or Internet Protocol (IP)
precedence because the number of different types packets
flows are classified by WRED but it is very limited that is
why still a better mechanism for congestion avoidance is
required [19].

11
M. K. Alam et al.
3. PROPOSED SCHEME
In this section we proposed three MAC layer schedulers to
analyze the randomness of the performance according to
different seed values because two schedulers named RED
and WRED mark the high dropping probability packets
randomly. Due to the randomness, the selection can be
good or bad. That is why, in the proposed paper,
modifying the seed values we can get the dissimilar
random values or positions. In this paper we analyze the
randomness of the performance of high speed WCN based
on various seed values for achieving different end to end
delay, and throughput of the multimedia applications
because every specific seed value generates different
specific random numbers or locations. In the sub-sections
we state the proposed system model and simulation
parameters.
3.1 Proposed System Model
In the proposed mechanism, We consider strict priority
scheduling mechanism for network layer for assigning the
priority on different ACs and three different MAC layer
scheduling schemes (FIFO, RED, WRED) to analyze the
randomness of the performance of WCN according to
different seed values. We set the beacon interval default
100ms which is fixed configured in Access Point (AP).
AODV routing protocol is employed to select the best path
to transfer the IMM traffic over the network. The random
waypoint mobility model is used to move the mobile
nodes as the real scenarios. Evaluating the effect of
randomness on different schemes of the proposed network
has been performed by modifying seed values. The packet
size 1560Bytes, and the traffic loads keep constant for
analyzing the randomness of the proposed network. Figure
2 shows the simulation animated veiw.
Fig. 2. The simulation scenario of the proposed network using Qualnet
simulator
3.2 Simulation Setup
The proposed work has been simulated by Qualnet 5.1
simulation tool. This simulation software provides scalable
simulation of different wireless networks. Qualnet offers
enormous simulation tools for comprehensive simulation
of wired and wireless networks. Using this software we
can simulate and emulate the networks to predict the
networks behavior, operation and management [20]. Table
III shows the different simulation parameters of the
proposed scheme.
Table 3: The Simulation Parameters
Simulator Used QualNET 5.1
Simulation Area 1000 X 1000
Simulation Time 100second
Total number of nodes 30
Propagation model Two-ray pathloss
Shadowing model Constant
MAC protocol IEEE802.11e
Radio Type IEEE802.11b
PHY data rate 11Mbps
Routing protocol AODV
Mobility model Random Waypoint
Application CBR
3.3 Performance Metrics
The performance evaluation of the proposed network has
been done by the different performance metric over
IEEE802.11e. These are throughput and end-to-end delay.
 Throughput (bits/sec) - the calculation of total
number of successful delivery packets divide by
the total amount of session between the first
packet sending start time and the last packet
received time.
 End-to-End Delay (sec) - represents the
combination of whole (source to destination)
travelling delay of packet.
4. SIMULATION RESULTS AND
DISCUSSION
In this section, we have simulated our proposed scheme
through WCN based on different seed value by qualnet 5.1
simulation tool. In this approach we measure the
performance of the proposed network through various
(End-to-End Delay and Throughput) performance metrics.
Following the sub-sections are analyses the simulation
result of the proposed scheme over WCN.

12
M. K. Alam et al.
4.1 Average End to End Delay and Throughput
Analysis
Figures 3 and 4 show the average end-to-end delay where,
the packet size and traffic load are constant always over
the network. After that, the proposed schedulers are having
different amount of average end-to-end delay and
throughput with respect to different seed values. The FIFO
scheduler is having the highest end-to-end delay because
there is no additional function to discard some packets
from the queue. In contrast, the RED and WRED can
remove the packets from the queue that packets are
marked randomly as dropping probability. Due to these
achieve low end-to-end delay. On the other hand, the FIFO
scheduler achieves better throughput compared to others
due to the least packets loss. Both Fig 3 and Fig 4 attain
different average end-to-end delay and throughput with
respect to various seed values because the randomness of
the proposed schedulers and networks. Finally, according
to this analysis we can say that, the effect of randomness
on the network might be good or bad, it depends on the
seed values and there is no guarantee to ensure QoS over
the network for IMM high priority traffics.
7
6
5
4
3
2
1
0
1 2 3 4 5
FIFO
RED
WRED
Seed Values
Average End-to-End Analysis (sec)
Fig. 3. Average End-to-End Delay Analysis
300000
250000
200000
150000
100000
50000
0
1 2 3 4 5
FIFO
RED
WRED
Seed Values
Average Throughput Analysis (bits/sec)
Fig. 4. Average Throughput Analysis
6. CONCLUSION
In this paper we analyze the randomness of the
performance of the proposed scheme based on different
seed values to justify the effect of the mentionable
parameter (seed value) on the proposed network
performance. After finished the simulation for every seed
value of the proposed scheme, we have proved that a great
change of the performance of our proposed scheme cause
of the randomness exist in the proposed network and the
variation of seed values. Thus, we can conclude that
analyzing the effect of randomness on the networks, seed
value is the very effective parameter. Evaluating the effect
of randomness of the proposed network on jitter and
packets delivery ratio will be analyzed in the future work.
References
[1] IEEE 802.11b Supplement to 802.11-1999, . Part 11:
Wireless LAN Medium Access Control (MAC) and Physical
Layer (PHY) Specifica-tions: Higher speed Physical Layer
(PHY) extension in the 2.4 GHz band, 1999.
[2] IEEE 802.11a, Supplement to IEEE 802.11-1999. Part 11:
Wireless LAN Medium Access Control (MAC) and Physical
Layer (PHY) Specifications: High-Speed Physical Layer in
the 5 GHz Band, 1999.
[3] Md. Khorshed Alam, Suhaimi Abdul Latif, Mosharrof H.
Masud, Farhat Anwar, " A Review on Scheduling Schemes
of High Speed Wireless Campus Network for Interactive
Multimedia Transmission", IEEE 11th Malaysia International
Conference on Communications (MICC), 2013.
[4] M. K. Alam, S. Abd Latif, M. H. Masud, M. Akter, F.
Anwar, J. S. Yalli, " An Analysis of Scheduling Scheme for
QoS Guaranteed Interactive Multimedia over High Speed
Wireless Campus Networks ", 1st International Conference
on Artificial Intelligence, Modelling and Simulation (AIMS),
2013.
[5] R. Dondera, C. Jia, V. Popescu, C. Nita-Rotaru, M. Dark, and
C. S. York, “Virtual classroom extension for effective
distance education.,” IEEE computer graphics and
applications, vol. 28, no. 1, pp. 64–74, 2008.
[6] K. Nisar, A. Amphawan, S. Hassan, and N. I. Sarkar, “A
comprehensive survey on scheduler for VoIP over WLAN,”
Journal of Network and Computer Applications, vol. 36, no.
2, pp. 933–948, Mar. 2013.
[7] IEEE, Part 11: Wireless LAN Medium Access Control
(MAC) and Physical Layer (PHY) specifications Amendment
8: Medium Access Control MAC) Quality of Service
Enhancements November 2005.
[8] Bello, Lucia Lo, Emanuele Toscano, and Salvatore Vittorio.
"A perspective on the IEEE 802.11 e Protocol for the Factory
Floor." (2010).
[9] Rojas, Andres, Philip Branch, and Grenville Armitage.
"Experimental validation of the random waypoint mobility
model through a real world mobility trace for large
geographical areas." In Proceedings of the 8th ACM
international symposium on Modeling, analysis and

13
M. K. Alam et al.
simulation of wireless and mobile systems, pp. 174-177.
ACM, 2005.
[10] C. Bettstetter, “Mobility Modeling in Wireless
Networks: Categor-ization, Smooth Movement, and Border
Effects,” ACM MobileComp. and Comm. Rev.,vol. 5, no. 3,
2001.
[11] T. Camp, J. Boleng, and V. Davies, “A Survey of
Mobility Models forAdHocNetworkResearch,” Wireless
Comm. & Mobile Computing (WCMC),vol. 2, no. 5, pp.
483-502, 2002.
[12] M. Musolesi,S. Hailes,and C. Mascolo. An ad hoc
mobility model founded on social network theory. In
Proceedings of the 7th ACM International Symposium on
Modeling, Analysis and Simulation of Wireless and Mobile
Systems. ACM Press,October 2004.
[13] Bettstetter, Christian, Giovanni Resta, and Paolo Santi.
"The node distribution of the random waypoint mobility
model for wireless ad hoc networks." Mobile Computing,
IEEE Transactions on 2, no. 3 (2003): 257-269.
[14] Weber, Reid M., and Michael M. Oberberger. "Gaming
system and method enabling player participation in selection
of seed for random number generator." U.S. Patent
Application 13/957,019, filed August 1, 2013.
[15] Kelly, John R., and Bruce McLoughlin. "Random
number generator seeding method and apparatus." U.S.
Patent 6,728,740, issued April 27, 2004.
[16] S.Rajeswari and Dr.Y.Venkataramani "Congestion
Control and QOS Improvement for AEERG protocol in
MANET ", International Journal on AdHoc Networking
Systems (IJANS) Vol. 2, No. 1, January 2012.
[17] James F. Kurose, Keith W. Ross " Computer
Networking : A top down approach (5th edition)",
http://engineering.free-book.biz/Computer-Networking-A-Top-
Down-Approach-5th-Edition-ID208.pdf, 2010.
[18] Hamadneh, Nabhan, David Murray, Michael Dixon,
and Peter Cole. "Dynamic Weight Parameter for the Random
Early Detection (RED) in TCP Networks." International
Journal of New Computer Architectures & their Applications
2, no. 2 (2012).
[19] Amara, Satish Kumar, and Michael Freed. "Policy-based
weighted random early detection method for avoiding
congestion in internet traffic." U.S. Patent 6,996,062, issued
February 7, 2006.
[20] Scalable Network Technologies, Qualnet User Manual
v 4.0., Culver City CA, www.qualnet.com,2010.

Impact of Randomness on MAC Layer Schedulers over High Speed Wireless Campus Network in IEEE802.11e

More Related Content

What's hot (19)

Viewers also liked (11)

Similar to Impact of Randomness on MAC Layer Schedulers over High Speed Wireless Campus Network in IEEE802.11e (20)

Recently uploaded (20)

Impact of Randomness on MAC Layer Schedulers over High Speed Wireless Campus Network in IEEE802.11e