International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395 -0056

Volume: 04 Issue: 02 | Feb -2017 www.irjet.net p-ISSN: 2395-0072

ANALYZING THE INTERACTION OF ASCENT WITH IEEE 802.11e MAC IN WIRELESS

SENSOR NETWORK
A.Sundar Raj1, S. Kavya2, K.G. Devi Priyanka3
Associate Professor,
E.G.S. Pillay Engineering College, Nagapattinam, India1
[email protected]
E.G.S. Pillay Engineering College, Nagapattinam, India2,3
[email protected], [email protected]

-------------------------------------------------------------***----------------------------------------------------------------
Abstract A wireless sensor network (WSN) consists of a [6]. There are many topology management schemes for
number of sensors distributed in the sensor field. These sensor network, like ASCENT, SPAN, BEES, STEM and GAF
sensors monitor physical or environmental conditions. In [1], [3], [9], [12]. They help the network to perform better
Adaptive Self-Configuring sEnsor Networks Topologies
by their management of controlling data packets, reducing
(ASCENT) the nodes can coordinate to exploit the redundancy
provided by the high density, so as to extend overall system
energy consumption and so on. We also surveyed many
lifetime. When the nodes detect high message loss, it requests congestion control schemes [8], [10], [11] to improve the
additional nodes in the region to join the network in order to topologies.
relay messages. This reduces nodes duty cycle if it detects In our work we have taken the ASCENT (Adaptive Self
high message losses due to collisions. Due to dense Configuring sEnsor Network Topologies) topology
deployment congestion becomes more common phenomenon management scheme [1], which is an adaptive network
from simple periodic traffic to unpredictable bursts of topology for mobile network. It can dynamically choose
messages triggered by external events. Congestion causes ACTIVE nodes to send messages successfully in the network.
huge packet loss and thus blocks reliable event perception.
In ASCENT the limitation is the packet loss due to
Congestion in wireless sensor networks causes packet loss,
and thus also leads to excessive energy consumption. So we
congestion. Hence the energy consumed will be higher and
need to control congestion in WSN in order to prolong system throughput will be less. To overcome this we use IEEE
lifetime. In order to reduce the congestion, we have 802.11e MAC layer and we analysis the performances like
implemented MAC layer 802.11e instead of CSMA. It is packet delivery ratio, system throughput and energy
observed that the proposed scheme performs well in terms of consumption and will be comparing ASCENT with CSMA.
packet delivery ratio, throughput and energy consumption. The paper is organised as follows, section II discusses the
related work in topology of wireless sensor network,
Keywords Wireless Sensor Networks, ASCENT, Topology, section III explains ASCENT, section IV explains the
MAC layer, Lifetime Improvement factor, NS2.
interaction of ASCENT with MAC 802.11e, in section V
I. INTRODUCTION results and analysis are discussed and conclusion is given in
the final section.
Wireless Sensor Network consists of large number of
sensor nodes deployed either uniformly or randomly in the II. RELATED WORK
field. These nodes will always be wasting energy as they C. Wang et al have proposed Upstream Congestion
will be in active state. Hence the main goal is to put some of Control in Wireless Sensor Networks through Cross-Layer
the sensor nodes in sleep state and some other nodes in Optimization [10]. In this congestion is classified as node
active state. Battery life, sensor update rates, and size are level congestion and link level congestion. Node level
all major design considerations in wireless sensor networks. congestion causes packet loss and queuing delay and this
Recent advances have resulted in the ability to integrate leads to retransmission, so consumes additional energy.
sensors, radio communications, and digital electronics into Link level congestion increases service time which leads to
a single integrated circuit package, which reduces size and decrease in both link utilization and overall throughput. In
complexity of the node. This capability is enabling networks this Priority-based congestion control protocol (PCCP) is
of very low cost sensors that are able to communicate with proposed. This employs packet based computation to
each other using low power wireless data routing protocols. optimize congestion control.
A wireless sensor network consists of a base station that Md. Mamun-Or-Rashid et al have proposed Reliable
can communicate with a number of wireless sensors. Data Event Detection and Congestion Avoidance in Wireless
is collected at the wireless sensor node, transmitted to the Sensor Networks [8]. Here congestion avoidance protocol
gateway directly or uses other wireless sensor nodes to [8] includes hierarchical medium access control (HMAC)
forward data to the gateway. The transmitted data is then and weighted round robin forwarding (WRRF) are
given to the system by the gateway connection. proposed. In HMAC node carrying higher amount of traffic
We have carried out an extensive survey on wireless gets more accesses than others. Therefore, downstream
sensor networks and topology management schemes [2], nodes obtain higher access to the medium than the

2017, IRJET | Impact Factor value: 5.181 | ISO 9001:2008 Certified Journal | Page 206
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395 -0056
Volume: 04 Issue: 02 | Feb -2017 www.irjet.net p-ISSN: 2395-0072

upstream nodes. This access pattern is controlled with local

values and is made load adaptive to cope up with various After Tt
application scenarios. In WRRF Downstream node allows Test Active
all of its upstream nodes to transmit their weighted-share
amount of packets. To avoid congestion, before Neighbors < NT
transmitting a packet each upstream node must be aware and Neighbors > NT
whether there is sufficient free buffer space at the Loss > LT Loss < LT
Loss < LT and help
downstream node. To implement this notion, they restrict
an upstream node from delivering packets when its After Tp
downstream node has not sufficient amount of free buffer Passive Sleep
space. This was achieved by their proposed source count After Ts
based weighted round robin forwarding (WRRF). The data Figure 1. ASCENT state transition
transmission in a network is termed as unreliable if the The state transition in ASCENT is shown in the Figure 1.
packet delivery ratio decreases to very low value so that the Each and every node in ascent first enters in the test state,
event cannot be detected reliably. Thus an efficient way is then it will check to go weather active or passive state
proposed to reduce congestion within the sensor network depends on the conditions. It goes to passive state if the
to ensure good delivery ratio for reliable event detection. number of active state nodes is greater than the threshold
Integrated effort of their proposed source count based value and loss is less than the threshold value. Otherwise it
HMAC and WRRF reduces packet drop due to collision and will go to active state after the time Tt. From passive state it
avoids packet drop due to buffer overflow and finally will go sleep or test state depending upon the following
achieves more delivery ratio which is good enough for condition, if the number of active state nodes is less than
reliable event perception. the threshold active state nodes and loss is greater than the
Chen et al have proposed a new technique called Span threshold loss or loss is less but help message is received
which is a topology control protocol that allows nodes that from the neighbour then it will go to test state, otherwise it
are not involved in a routing backbone to sleep for will go to sleep state after the time Tp. From sleep state it
extended periods of time [3]. In this topology, certain nodes will go to passive state after the time Ts. By this way the
assign themselves the position of coordinator. These required number of active state nodes will be maintained in
coordinator nodes are chosen to form backbone network, the network for the reliable transfer of data.
so that the capacity of the backbone approaches the When a node sends message to the sink, the sink may be
potential capacity of the complete network. Periodically, at the border of radio range so message loss will occur. In
nodes that have not assigned themselves the coordinator this case the sink starts sending help messages to the
role initiate a procedure to decide if they should become a neighbours that are in listen-only (passive) mode. When the
coordinator. The criterion for this transition is if the node receives help message it may join the network by
minimum distance between any two of the nodes becoming active state from the conditions shown in the
neighbours exceeds three hops. Figure 1. When a node decides to join the network, it
signals the existence of a new active neighbour to other
III. ASCENT
passive neighbours by sending a neighbour announcement
Adaptive Self Configuring sEnsor Network Topology is a message. After the required number of active state nodes
wireless sensor topology management scheme that has four are obtained the delivery of data from source to sink is now
states of nodes namely ACTIVE, PASSIVE, TEST and SLEEP. more reliable.
In the active state transceivers will be active all the time Notation and definition:
and participates in the communication. An enough number 1. Tt Test timer, time required to check the
of active state nodes will be required to transmit the data condition specified in Figure 1.
successfully to the sink. In passive and test states, the nodes 2. NT Number of Neighbour threshold, we assume
receiver will be turned on. It will listen for the messages this as 4.
flooded in the network and these states will seek to change 3. LT Maximum amount of data loss it can tolerate
as active states whenever needed to participate in the depends on application.
network. Sleep state is the energy saving mode in which 4. Ts Sleep timer.
transceivers will be turned off. 5. Tp Passive timer
When a node detects an event, it will be routed through For conserving energy the node must remain in sleep
active state nodes. In case if a node detects message loss it state, but if Ts is large then that node cant participate in the
will send help messages to the neighbours. Node which is in network when required. If the node remains in the passive
passive state receives these messages and checks for the state it can make transition to active state whenever
conditions to become active node if needed. required, but if Tp is large energy consumed will be high. So,
there must be a tradeoff between Tp and Ts for proper
network operation.

2017, IRJET | Impact Factor value: 5.181 | ISO 9001:2008 Certified Journal | Page 207
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395 -0056
Volume: 04 Issue: 02 | Feb -2017 www.irjet.net p-ISSN: 2395-0072

The probability of nodes in passive or sleep state is given We have implemented ASCENT and then combined
by ASCENT with IEEE 802.11e and thereby performance is
analysed by comparing ASCENT with CSMA. The ASCENT
P(passive) (1)
1 with IEEE 802.11e scheme has better performance in terms
P(sleep) 1 P(passive) (2) of packet delivery ratio, throughput and energy
conservation when compared to ASCENT with CSMA MAC
The probability of x nodes in the passive state is given by layer.
n
1 x 1
p( x ) 1 . (3) A. Simulation Environment
1 1 Simulation is carried out using Network simulator (NS2).
Here is the ratio of passive time Tp to the sleep time Ts In our simulation we have considered random deployment
and n is number of nodes. of nodes with mobility. We have taken the number of nodes
By taking x=1 and x=2 we can find the minimum value of as 20, 40, 60, 80 and 100 and plotted against various
as 1 and 2 respectively. The solution is given by parameters like packet delivery ratio, throughput and
1
log(1p t ) energy consumption. Their performance evaluation is
1 10 n 1 (4) discussed below. We have taken initial energy for each
1
log(1 p t ) node as 1000 J.
2 10 n
1 1 (5)
Where pt is minimum probability threshold.
With these values we can set the sleep time (Ts) and B. Dependency of value on Pt
passive time (Tp).
1
IV. INTERACTION OF ASCENT WITH IEEE 802.11 x=1,Pt=99%
0.9
In our proposed work we have analysed the interaction x=1,Pt=95%
of ASCENT with IEEE 802.11e MAC layer. This MAC layer is 0.8 x=2,Pt=99%
x=2,Pt=95%
specially designed to provide quality of service in a network 0.7
[5]. To achieve this, a new coordination function is used
0.6
namely hybrid coordination function (HCF). The HCF
Gamma

combines the functions of DCF and PCF with some 0.5

enhanced QoS-specific mechanisms and frame subtypes to 0.4
allow a uniform set of frame to exchange sequences during
both the contention period (CP) and contention free period 0.3

(CFP). The HCF uses both contention-based channel access 0.2

methods, called the enhanced distributed channel access
0.1
(EDCA) mechanism and controlled channel access, referred
to as the HCF controlled channel access (HCCA) mechanism. 0
0 10 20 30 40 50 60 70 80 90
The EDCA mechanism provides differentiated and Number of nodes
distributed access to the wireless medium for QSTAs
(quality of service station) using eight different user Figure 2. Dependency of value on Pt
priorities (UP). The EDCA mechanism defines four access From Figure 2 it is inferred that for higher value of Pt the
categories that provide support for the delivery of traffic value remains high for corresponding node density, for
with UPs at the QSTAs. lower values of Pt the value are low. Thus the estimation
Higher priority traffic has a more chance of being sent of value has more dependency on Pt values.
than low priority traffic, a station with high priority traffic
waits for less time before it sends its packet, than a station C. Packet delivery ratio
with low priority traffic which as to wait little more [5].
This is achieved by using a shorter contention window (CW)
and shorter arbitration inter-frame space (AIFS) for higher
priority packets. EDCA also provides contention-free access
to the channel to a period called a Transmit Opportunity
(TXOP). A TXOP is a time interval during which a station
can send many frames without congestion. The frame has to
be fragmented, when the frame size is very large that cant
be transmitted in a single TXOP.

V. RESULTS AND ANALYSIS

2017, IRJET | Impact Factor value: 5.181 | ISO 9001:2008 Certified Journal | Page 208
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395 -0056
Volume: 04 Issue: 02 | Feb -2017 www.irjet.net p-ISSN: 2395-0072

Packet Delivery Ratio

100
E. Energy Consumption
90
Energy consumed
CSMA 1000
80
IEEE 802.11e
Packet Delivery ratio %

70 900

60
800
CSMA
50
IEEE 802.11e

Consumed energy
700
40

30 600

20 500

10
400
0
20 40 60 80 100 120
300
Number of nodes

Figure 3. Packet Delivery Ratio 200

20 40 60 80 100
Number of nodes
The packet delivery ratio for different number of nodes is
studied and plotted for ASCENT with IEEE 802.11e MAC Figure 5. Energy consumption.
and with CSMA MAC as shown in Figure 3. It is observed
that in both the cases the packet delivery ratio drops as the
number of nodes increases. Since the collision is minimised Figure 5 shows energy consumed by ASCENT with IEEE
in ASCENT with IEEE 802.11e, packet delivery ratio 802.11e and ASCENT with IEEE CSMA MAC layer. The
increases substantially. energy consumed by ASCENT with IEEE 802.11e is much
lesser than energy consumed by ASCENT with CSMA MAC
layer. From this we can infer that ASCENT with IEEE
802.11e consumes very less energy.

VI. CONCLUSIONS
In this paper we have implemented ASCENT with CSMA.
D. Throughput We have integrated ASCENT with IEEE 802.11e MAC layer.
Throughtput of the system We analysed the parameters namely packet delivery ratio,
1500 system throughput and energy consumption for both CSMA
and IEEE 802.11e MAC layers. It is inferred that ASCENT
CSMA with IEEE 802.11e MAC layer provides higher packet
IEEE 802.11e
delivery ratio, increased system throughput and lower
1000 energy consumption when compared to ASCENT with
Throughput (B/s)

CSMA MAC layer. Thus ASCENT with IEEE 802.11e

performs better in wireless sensor network.

REFERENCES
500
[1] Alberto Cerpa and Deborah Esttrin, ASCENT: Adaptive Self-
Configuring sEnsor Networks Topologies, IEEE Transactions on
mobile computing, Vol.3, July-September 2004.
[2] Akyildiz. I.F, Su. W, Sankarasubramaniam. Y, Cayirci. E, Wireless
sensor networks: a survey, 2002 Published by Elsevier Science B.V,
0 Computer Networks 38 (2002) pp.392-422.
20 40 60 80 100 120 [3] Chen. B, Jamieson. K, Balakrishnan. H, Morris.R (2002), Span: An
Number of nodes Energy-Efficient Coordination Algorithm for Topology Maintenance
in Ad Hoc Wireless Networks Springer, 8(5).
Figure 4. Throughput [4] Curt Schurgers, Vlasios Tsiatsis, Mani B. Srivastava, STEM:
Topology Management for Energy Efficient Sensor Networks,
Throughput of the network is analysed and it is shown in IEEEAC paper #260, Updated Sept 24, 2001.
Figure 4. The throughput of the network is higher in [5] IEEE Std 802.11e-2005
ASCENT with IEEE 802.11e when compared with ASCENT [6] Mo Li, Baijian Yang: A Survey on Topology issues in Wireless Sensor
with CSMA. From this it is inferred that ASCENT with IEEE Network In Proceedings of ICWN, Las Vegas, Nevada, USA, June
2006, pp.503 - 509.
802.11e MAC provides higher system capacity.

2017, IRJET | Impact Factor value: 5.181 | ISO 9001:2008 Certified Journal | Page 209
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395 -0056
Volume: 04 Issue: 02 | Feb -2017 www.irjet.net p-ISSN: 2395-0072

[7] Miguel A. Labrador and Pedro M. Wightman, Topology Control in [10] Wang. C, Li. B, Sohraby. K, Daneshmand. M and Hu. Y, Upstream
Wireless Sensor Networks, Springer April 2009 Congestion Control in Wireless Sensor Networks Through Cross-
[8] Mamun-Or-Rashid. Md, Muhammad Mahbub Alam, Abdur Razzaque. Layer Optimization, IEEE Journal On Selected Areas In
Md, and Choong Seon Hong, Reliable Event Detection and Communications, Vol. 25, No. 4, May 2007
Congestion Avoidance in Wireless Sensor Networks,HPCC 2007, [11] Xiang-Yang Li, Kousha Moaveni-Nejad, Wen-Zhan Song, Wei-Zhao
LNCS 4782, pp521-532, 2007. Wang, Interference-Aware Topology Control for Wireless Sensor
[9] Rong Yu, Zhi Sun, and Shunliang Mei (2007), Scalable Topology and Networks, IEEE SECON 2005 proceedings.
Energy Management in Wireless Sensor Networks, in the [12] Y. Xu, J. Heidemann, D. Estrin, Geography-informed energy
proceedings of the IEEE WNCA. conservation for ad hoc routing, MobiCom 2001, Rome, Italy, pp.
70-84, July 2001.

2017, IRJET | Impact Factor value: 5.181 | ISO 9001:2008 Certified Journal | Page 210