Bahria University Journal of Information & Communication Technologies Vol. 9, Issue 1, June 2016
Energy and Priority Aware Scheduling Scheme
for Wireless Sensor Networks
Saira Arif, Gulistan Raja
Abstract — Wireless sensor networks (WSNs) are playing a
vital role in automation, sensing and commercial applications.
Key issues in WSNs are to satisfy Quality of service (QoS)
parameters and to conserve energy. Some of the sensor nodes in
WSN act as a single communication link to connect one group
of sensor nodes to other in multi-hop environment. If nodes
acting as a single communication link become dead due to low
energy then this can cause network failure and nodes causing
failure are called bottleneck nodes. This paper proposes an
Energy and Priority Aware Scheduling (EPAS) scheme for
bottleneck node avoidance. Every node has three priority
queues to place packets of different priorities. Scheduling of
packets is performed based on energy level threshold and
priority level marked. Proposed scheme avoids bottleneck
problem by scheduling packets of high priority based on energy
level thresholds, if energy is critically low then low priority
packets are dropped but sensor node still keep transmitting high
priority packets in order to keep network alive. Proposed
scheme is implemented using Network Simulator-2 (NS-2).
Simulation results show that the scheme achieves 96% packet
delivery ratio for highest priority packets and gives 52%
improvement in packet delivery ratio, 0.309 Kbits/sec improved
throughput and 5.16s less delay as compared to priority
queuing.
Index Terms — QoS: Quality of service, WSN, NS-2
I. INTRODUCTION
Wireless Sensor Networks (WSN) are group of sensor
devices which collaborate to monitor any environmental or
physical phenomenon such as temperature, humidity level,
pressure while working at diverse locations. Considering that
most of wireless sensor nodes are battery operated, energy
conservation is one of biggest challenges as sensors once
deployed in an environment remain there for years and it is
difficult to recharge or replace batteries. Besides overall
energy consumption, Quality of Service (QoS) differentiation
and reliability are also important features demanded by
today’s WSN environment [2]. QoS is capacity of network to
fulfill certain requirements of a user or an application;
requirements can be minimum bandwidth, packet delay,
packet loss etc. QoS differentiation for WSNs can be defined
as different sensors require different level of QoS to perform
different tasks in a system. For example, in surveillance
system status of an anomaly or intruder must be updated
within few seconds of detection. Similarly status of fire alarm
is more prior to water level in a home automation system.
Thus applications requiring QoS are gaining more and more
importance in field of WSNs. Many researchers designed
S. Arif, and G. Raja, Faculty of Electronics and Electrical Engineering, UET
Taxila. Email: [email protected]. Manuscript received October 10,
2015; revised on February 25 and April 29, 2016; accepted on May 20,
2016.
Page 36
routing protocols, Medium Access Control (MAC) protocols
and scheduling schemes to provide energy efficiency and
ensure QoS to WSNs. Scheduling schemes play a vital role at
sensor nodes since they guarantee delivery of different data
types based on their priority and ensure minimum delay for
each type of packet. Literature review shows that very fewer
studies exist on packet scheduling that schedule data
processing at sensor nodes and also reduce energy
consumption. In fact most operating systems for WSNs use
First in First out (FIFO) scheduling which process packets as
per arrival time i.e. packets received first are transmitted first.
However packets at sensor nodes must be delivered
considering different constraints like deadline, delay, queue
length and priority levels. Most of scheduling schemes
available in literature are not designed considering dynamic
requirements of WSNs.
In this paper a new energy and QoS aware scheduling
scheme named Energy and Priority Aware Scheduling
(EPAS) scheme is proposed. It uses priority marking to
provide QoS among nodes. Scheduling is performed
considering energy of node and priority level marked. Rest of
paper is organized as follows. In section II an overview of
related research work is given. Section III explains proposed
EPAS scheme in detail while section IV gives performance
evaluation of proposed scheme using Network Simulator
(NS-2). In the end section V concludes the paper.
II. RELATED WORK
This section gives an overview of literature related to
Quality of Service (QoS) aware MAC protocols, scheduling
schemes and bottleneck nodes. N. Saxena et al. presented
idea of a MAC protocol designed specifically for multimedia
transmission over WSNs [4]. Aim is to conserve energy and
provide QoS support. It varies duty cycle and Congestion
Window (CW) based on network parameters such as
transmission failure rate and priority of traffic class.
Differentiation of QoS among different traffic classes is
provided by varying CW sizes using different multiplicative
coefficients among traffic classes. There is no centralized
control so early sleeping as well as idle channel listening are
main drawbacks. Also protocol cause high latencies for
packets of low priority type.
Priority Reservation Medium Access Control (PRMAC) is a MAC protocol which assigns priorities to sensors
based on the event monitored by that sensor node e.g. in a
home automation system a sensor for fire alarm is more
important than a sensor for water level [5]. Different
priorities are assigned to sensors for every event under
observation by sensor node and differentiation between
events is provided by changing Congestion Window (CW)
size and varying Inter Frame Space (IFS) per event. Sender
ISSN – 1999-4974
Bahria University Journal of Information & Communication Technologies Vol. 9, Issue 1, June 2016
node communicates a small duration pulse instead of using
RTS-CTS messages for medium reservation. Later M.A
Yigitel et al. proposed Q-MAC [6], protocol is designed to
provide QoS among nodes and energy conservation for multihop networks. It is a scheduling based protocol and schedule
priorities of packets considering their hop count, remaining
energy and node’s queuing load. Q-MAC uses intra-node
scheduling to choose next packet to service among different
priority queues and medium access among neighboring nodes
is provided by inter-node scheduling. Packets are queued at
their respective nodes, as per their priority in FIFO manner
and then from respective nodes highest priority packet is
transmitted towards receiver using different algorithms for
intra-node and inter node scheduling. However computing
transmission urgency for any sensor is complex and also
leads to higher latencies.
Congestion avoiding algorithms for multi-hop wireless
environment are proposed to work at different layers of OSI
model [ 9, 10, 11]. B. Ngo et al. proposed a scheme to address
congestion at relay nodes by balancing its relaying activities.
[12]. Relay node tries to balance its incoming and outgoing
traffic. If receiving rate is too high relay node starts dropping
RTS frames. Whenever there is an imbalance between
incoming and outgoing traffic, relay node drops RTS frames,
it entertains next packet only when buffered queue is empty.
This scheme achieves 100% throughput increase as
compared to IEEE 802.11 RTS/CTS strategy. High dropping
rate of Channel request frames provides feedback to nodes at
every hop. Thus packet sending rate of nodes is limited by
relay nodes leading to stabilize the system.
Later R. K Tripathi et al. propose an algorithm for
bottleneck nodes detection and critical level of a node [13].
In this algorithm nodes are divided into two types, root nodes
who initiate flooding and those which are used to forward
signals of root node. Root nodes are determined from number
of neighboring nodes, a node having least number of
neighbors is considered a root node. Reason of this criterion
is nodes having least number of neighbors have more chances
of being at border of network. Thus being at border of
network can act as bottleneck. If any node receive flooding
messages from other than one root node in that case it is
critical node of network, i-e it can be considered bottleneck.
Criticality of a node increases as number of root nodes
connected to it increases. Thus in this way bottleneck node
detection is done in this algorithm.
Later on authors reported the concept of a congestion
control scheme for WSNs named ECODA (Enhanced
Congestion Detection and Avoidance) [14]. It is designed to
operate at both MAC and network layer and works for
multiple traffic classes. Scheme proposed uses cross layer
design, detects congestion using buffer and weighted buffer
difference. Packet transfer is done using flexible queue
scheduler, packets are assigned static and dynamic priorities
and packet scheduler selects next packet as per priority.
Queuing strategy used commonly for WSNs is FIFO, it
schedules packets in their order of arrival i-e packets which
come first are served first but it does not consider constraints
like deadline, delay or priority. Scheduling packets in a
priority based method involves use of Priority Queuing (PQ).
Page 37
It can use single queue architecture with high priority packets
placed at top of queue or multiple queues can be used for
queuing packets of different priority in different queues.
Every queue is a FIFO queue. High priority values are
assigned to control packets mostly while data packets are
assigned low priority values. Packets from high priority
queues are served first. If congestion occurs low priority
packets are dropped first while there is no fair distribution of
resources for all types of packets [15]. In fair queuing (FQ),
data packets are identified as flows and every flow is stored
in a particular queue. It uses round robin algorithm to ensure
equal sharing of resources among flows. Weighted Fair
Queuing (WFQ) is a combination of PQ and FQ. Just like FQ
method, all queues are served to avoid bandwidth starvation,
but weights are assigned to queues so a queue with high
weight use network for longer duration or it is served first.
Thus assigning weight is just like giving different priorities
to queues. Thus classification of packets is done and then they
are assigned different queues accordingly [15] but these
scheduling schemes are not designed considering dynamic
requirements of WSNs. Nidal Nasser et al. proposed
Dynamic Multi Priority (DMP) scheduling scheme for WSNs
,it differentiates packets into real and non-real time data [16].
It uses multiple variable length queues, real time packets
which are of highest priority so placed in high priority queue
while non-real time packets are placed in low priority queues
considering certain threshold of time. Real time packets can
preempt transmission of low priority packets. It improves
QoS by delivering real packets fast and reducing delay. A
performance analysis of DMP is presented by S. Bansode et
al. using NS-2 [22]. Overview of literature suggest that most
of QoS related research work assigns priorities to nodes either
by adjusting CW size or varying duty cycle but they have
drawback like complex cost function, high latencies and
starvation of low priority packets. Scheduling schemes are
designed considering traffic classes, single queue and
multiple queue architecture, priority levels and deadlines.
There is research space to design a scheduling scheme
considering dynamic requirements of WSNs to avoid
bottleneck node problem.
III. PROPOSED SCHEDULING SCHEME
Proposed Energy and Priority Aware (EPAS)
Scheduling scheme provides QoS to wireless sensor nodes
while efficiently using energy resources. It uses prioritization
to provide QoS, multi-queue architecture and energy aware
decision making to avoid bottleneck problem. System model
for proposed scheduling scheme consists of a group of
wireless sensor nodes connected by a single communication
link named bottleneck link as shown in Fig. 1. Sensor nodes
within group can communicate from one end to other as long
as bottleneck link is active. If this link is broken or down
because of low energy then network will be partitioned and
there will be no communication between different parts of
network. Let sensor nodes have different data sending rates
and there is no priority based classification of packets and
nodes. If bottleneck nodes use simple scheduling techniques
like first come first serve then a low priority sensor node with
ISSN – 1999-4974
Bahria University Journal of Information & Communication Technologies Vol. 9, Issue 1, June 2016
high data sending rate may occupy channel for a long
duration causing starvation for other sensor nodes and
bottleneck node may drain its energy sending low priority
packets. Thus there is a need of priority based classification
of data packets and prioritize scheduling to address the
problem of bottleneck nodes.
used as number of priority levels of packets is three,
scheduler checks priority level marked on packet and assigns
every packet its particular queue. The proposed scheme uses
following notation for all the three queues.



Fig. 1 System Model for Proposed Scheme.
Proposed scheduling scheme consists of following phases
1) Priority marking
2) Priority queue assignment
3) Energy and priority aware scheduling
A. Priority Marking
Every node mark data packets with static priorities
depending upon importance of task assigned to them. Every
packet is assigned a priority level and a priority queue as per
priority of packet, in this scheme we have three priority levels
for packets
1) High
2) Mid
3) Low
Every packet is marked with a priority level and when it
is received by a node it is assigned a particular queue as per
its priority. In this scheme packets are assigned three priority
levels, packets having highest priority are marked as P1, then
comparatively low priority packets are marked with P2 and
lowest priority packets are given priority of P3. Table I gives
details of priority assignment in tabular form. Main
advantage is that QoS differentiation among packets is
achieved by marking priority levels.
Table 1. Priority Assignment Table
Priority value
Packet status
P1
Highest priority packet
P2
Mid level Priority
P3
Low priority
B. Priority Queue Assignment
In this scheme we have used approach of multiple
priority queues. Typical PQ uses a single queue with high
priority packets placed at the top of queue but in this scheme
every node has three FIFO queues. Three priority queues are
Page 38
HPQ-1 represents queue containing packets of highest
priority i.e. packet of type P1.
HPQ-2 represents queue with packets of mid-level
priority i.e. packets of type P2.
LPQ denotes queue containing lowest priority packets
i.e. packets of type P3.
C. Energy and Priority Aware Scheduling
It is most important phase of proposed scheduling
scheme as packets are differentiated into different types as
per importance of task and energy level of node. This scheme
uses three energy levels as decision thresholds. Initial energy
of node is compared with remaining energy of node. Every
node whenever receives a packet it checks packet’s priority
level marked in its header and its own energy level. We have
used three energy thresholds.
1) Threshold 1: Remaining Energy of node > 70% of initial
energy
2) Threshold 2:Remaining Energy of node > (30% and <
=70% of initial energy)
3) Threshold 3: Remaining Energy of node < 30% of initial
energy.
Scheduling of packets is performed considering these
thresholds and priority levels, detailed process is explained in
next section.
D. Working Principle
In this section we explain working of proposed EPAS
scheduling scheme. All nodes assign priority levels to
packets as per importance of task assigned as discussed in
priority assignment block. When a bottleneck node receives
any packet it examines priority level of packet marked in its
header and then it checks its own remaining energy. If energy
level of node is greater than 70% of initial energy then it will
allow all priority packets i-e P1, P2 and P3 to transmit with
packets of priority level P1 are transmitted first, then P2 and
at the end packets of priority P3 are transmitted. If energy
level is less than 70% but greater than 30% of initial energy
then only packets of priority P1 and P2 will pass through and
remaining packets are queued. If high priority queues (HPQ1 and HPQ-2) are empty then packets from low priority queue
are allowed to pass through. If energy of node becomes less
than above two thresholds and drops below 30% of initial
energy then packets of highest priority P1 are transmitted
only while low priority packets are queued. If high priority
queue is empty then packets from low priority queues are
transmitted but if high priority queue HPQ-1 is not empty
then low priority packets are dropped in case of critically low
energy state i-e less than 30%. Thus using this scheme
bottleneck problem is avoided as even if energy of node is
critically low still it can pass high priority packets so that
network is alive. Proposed scheme improves overall network
life time, uses energy intelligently and provides QoS
ISSN – 1999-4974
Bahria University Journal of Information & Communication Technologies Vol. 9, Issue 1, June 2016
differentiation. Proposed scheduling scheme uses algorithm
given in Table II and parameters for algorithm are:
I.E = initial energy of node
E = current energy of node
Priority levels for packets i-e P1, P2 and P3
HPQ-1 queue = High Priority Queue for packet of priority
P1.
HPQ-2 queue = Mid Priority Queue for packet of priority P2.
Results are evaluated for packets of different priorities
separately. We have used cbrgen tool to generate traffic
connection pattern files while for scenario generation
Random Way Point Mobility model (RWP) [20] available in
NS-2 is used. For simulation three connection pattern files
with maximum of 10, 15 and 20 nodes connected also called
traffic flows whereas scenario consists of four mobility
pattern files with varied pause time for mobile nodes. Details
are given in Table IV.
Table 2. Algorithm of Proposed EPAS scheme
Table 4. Mobility Pattern files
Proposed Algorithm
Initialize priority levels to packets
P1 for high priority
P2 for mid priority
P3 for low priority
When a bottleneck node receives a packet
{ check priority level of packet && update energy status
if ( E > 0.7*I.E)
allow all priority packets to pass
else if ( E >= 0.3*I.E & < 0.7* I.E )
{ if (HPQ-1 || HPQ-2 queue is not empty)
transmit packets of priority P1 and P2.
else transmit P3 packets.
}
else if (HPQ-1 queue is not empty)
Allow packets of priority P1 only.
else
Allow packets of priority P2 or P3.
end if }
IV. PERFORMANCE ANALYSIS
Performance analysis of proposed scheduling scheme in
terms of data delivery probability, throughput and delay is
performed in this section. Simulation of proposed scheme has
been performed using NS-2.35 [19]. It is evaluated for
metrics of packet delivery ratio, throughput and delay.
A. Simulation Setup
Simulation setup consists of 25 mobile sensor nodes
placed randomly in an area of 500*500. CBR traffic is used
to send data over nodes. Every sensor node uses three priority
queues to place packets of priority level P1, P2 and P3
respectively. Simulation parameters used for proposed EPAS
scheme are given in detail in Table 3.
Table 3. Simulation Parameters
Simulation Parameter
Network Simulator
Routing Protocols
Size of Network
Number of sensors
Queue
Ifq-length
Packet Size
Data Type
Mac protocol
Initial Energy
Simulation Time
Propagation
Page 39
Value
NS-2.35
AODV
500m*500m
25
Priority and Normal
50 packets
512
CBR
802.11
10-100 Joule
100s
Mobility pattern
Description
mob-00
Mobility pattern file with zero pause time i-e
continuously moving
mob-03
Mobility pattern with pause time = 3 sec
mob-07
Mobility pattern with pause time = 7 sec
mob-10
Mobility pattern with pause time = 10 sec
B. Packet Delivery Ratio (PDR)
Packet delivery ratio (PDR) is a measure of successful
transmission rate and it is an important parameter to
determine effectiveness of system. PDR of packets is denoted
by P1-EPAS, P2-EPAS and P3-EPAS for high priority
packets, mid level priority packets and low level priority
packets. PDR is computed experimentally for all mobility
scenarios mob-00, mob-03, mob-07 and mob-10 separately
as shown in Fig. 2. PDR is calculated as a function of number
of connections also called traffic flows. Results show that
packets of highest priority P1 in proposed EPAS scheme
achieves maximum data delivery for all mobility scenarios
while PDR is low for low priority packets. Comparison of
proposed scheme results with PQ [21] for all mobility
scenarios show that proposed scheme performs better as
compared to PQ. Optimum results are obtained for mobility
scenario mob-07 with pause time equal to 7 sec because
movement of sensor nodes is moderate. In case of mob-00
nodes are continuously moving so chances of collisions are
high as shown in Fig. 2 (a). Overall results show proposed
scheme achieves maximum PDR for packets of priority P1
i.e. 95-96 % while low priority packets have comparatively
low data delivery ratio. Average results for all mobility
patterns are given in tabulated form in Table V. Comparing
values of P1 type packets in proposed scheme and PQ for
mob-07 in Table V shows that proposed scheme achieves
52% improvement in PDR.
Table 5. Average PDR of Proposed Scheme and PQ for all
mobility patterns
Priority Queue Type mob-00
mob-03
mob-07
mob-10
P1-EPAS
92.3
88.33
95
93.3
P2-EPAS
77
85
79
82.6
P3-EPAS
71
66
55
69
PQ
25
31
43
22
Two Ray Ground
ISSN – 1999-4974
Bahria University Journal of Information & Communication Technologies Vol. 9, Issue 1, June 2016
(a)
PDR as a function of No. of connections for mob-00.
(c) PDR as a function of No. of connections for mob-07.
(b) PDR as a function of No. of connections for mob-03.
(d) PDR as a function of No. of connections for mob-07.
Fig. 2 Packet delivery ratio for all mobility patterns.
C. Throughput
Throughput depends on number of packets received for a
particular node, performance plot of throughput versus no. of
connections for all mobility scenarios is given in Fig. 3.
Highest priority packets are denoted by P1-EPAS, mid level
priority packets are denoted by P2-EPAS and lowest priority
packets are denoted by P3-EPAS. Throughput graphs show
that proposed queuing scheme achieve different throughput
for all types of priority packets. With an increase in number
of connections also called traffic flow patterns throughput
increases for proposed scheme. Packets of type P1 have
maximum throughput as compared to low priority packets and
PQ in all mobility scenarios. Mobility pattern mob-07
achieves best results for throughput as compared to other
mobility scenarios while mob-03 scenario performs better as
compared to mob-00 because in mob-00 sensor nodes are
continuously moving so chances of collisions are more.
Overall results show that proposed scheduling scheme
performs better as compared to PQ scheduling while highest
priority packets P1 achieve maximum throughput.
D. Average Delay
Table 4 shows average delay computed for proposed
scheduling scheme and PQ for all mobility patterns. It shows
that proposed scheme gives minimum delay of 0.01sec for
mob-00 and maximum value of 2.5sec for mob-10 and an
average of 1.83sec for packets of type P1 i.e. highest priority
packets. PQ gives a minimum delay of 3.99sec for mob-00
Page 40
and maximum value of 8.6sec for mob-10 with 6.99sec
average delay for all mobility patterns. Comparison of
average values of proposed scheduling scheme and PQ shows
it gives 5.16sec decrease in delay.
Table 6. Average Delay (s) Vs. Mobility Pattern
Queue type
mob-00
mob-03
mob-07
mob-10
P1-EPAS
0.01
1.82
3
2.5
P2-EPAS
7.3
6.06
6.79
7.5
P3-EPAS
8.23
4.09
7.5
6.2
PQ
3.99
7.81
7.58
8.6
In Fig. 4 simulation results are shown in graphical form,
it is a plot of average value of delay for all mobility patterns
where each mobility pattern is simulated by varying no. of
connections. Fig. 4 shows that average delay is minimum for
highest priority packets i.e. P1 while for low priority packets
delay is more for proposed scheduling scheme. PQ gives low
value of delay in case of mob-00 while for remaining
scenarios it gives high delay values. Thus overall simulation
results show that proposed schemes gives better results as
compared to PQ scheduling in terms of PDR, throughput and
delay and avoids bottleneck problem by sending highest
priority packets in case of extreme low battery conditions.
ISSN – 1999-4974
Bahria University Journal of Information & Communication Technologies Vol. 9, Issue 1, June 2016
(a)
Throughput as a function of No. of connections for mob-00.
(c) Throughput as a function of No. of connections for mob-07.
(b) Throughput as a function of No. of connections for mob-03.
(d) Throughput as a function of No. of connections for mob-10.
Fig. 3 Throughput for all mobility patterns
Fig. 4 Average Delay (s) as a function of mobility
patterns
Proposed scheme has been implemented using NS-2.
Performance is evaluated based on various metrics including
PDR, throughput and delay under different mobility
scenarios. Experimental results show that proposed scheme
gives PDR of 96%, throughput of 2.785 Kbits/sec and
minimum delay of 0.01 sec for highest priority packets i.e. of
type P1. PQ achieves PDR of 43 %, throughput of 2.476 Kbits
and minimum delay of 3.99 sec for same simulation scenarios.
Thus the proposed scheme out performs PQ in terms of packet
delivery ratio, throughput and delay. Through detailed
simulation analysis, it can be concluded that the proposed
scheme gives promising results in a dense wireless
deployment, minimize chances of bottleneck node
development and improves overall network performance. The
proposed scheme implementation for Destination Sequenced
Distance Vector (DSDV) and Dynamic Source Routing
(DSR) can be investigated as a future work. Implementation
as a real time scenario is another milestone envisaged in near
future.
V. CONCLUSION AND FUTURE WORK
REFERENCES
In this paper, a new scheduling scheme considering
energy and priority marking for bottle neck nodes is proposed
for WSNs. It is proposed as a solution to bottleneck node
problem for WSNs. Proposed scheduling scheme is designed
to satisfy QoS needs of a network in terms of priority marking
while energy constraints are satisfied using threshold. It
avoids bottleneck node problem by dropping low priority
packets if energy is critically low but transmission of high
priority packets is maintained. This is the main advantage of
this scheme that it utilizes energy resources intelligently.
Page 41
[1] Y. Jennifer, M. Biswanth and G. Dipark (2008, August).
Wireless Sensor Network: A Survey Computer Network , vol.
52(12). pp. 2292-2330.
[2] J. N. Al-karaki and A. E. Kamal (2004, December). Routing
Techniques In Wireless Sensor Networks: A Survey. IEEE
Wireless Communications,vol.11(6). pp. 6-28.
[3] A. Bachir, M. Dohler, T. Watteyne and Kin K. Leung (2010,
April), MAC Essentials for Wireless Sensor Networks. IEEE
Communications Surveys & Tutorials , vol. 12(2).pp.222-248.
ISSN – 1999-4974
Bahria University Journal of Information & Communication Technologies Vol. 9, Issue 1, June 2016
[4] N. Saxena, A. Roy and J. Shin (2008, September). Dynamic
Duty Cycle and Adaptive Contention Window Based QoSMAC Protocol for Wireless Multimedia Sensor Networks.
Computer Networks. vol. 52 (13). pp. 2532–2542.
[5] A. Firoze, L. Ju, L. Kwong, “PR-MAC a priority reservation
MAC protocol for Wireless Sensor Networks”, in Proc. of
ICEE, Lahore, 2007 pp. 1–6.
[6] M. Aykut Yigitel, O.D Incel and C. Ersoy. (2011, June). QoSaware MAC protocols for Wireless Sensor Networks: A survey.
Computer Networks, Vol.55 (8). pp. 1982-2004.
[7] Y. Liu, I. Elhanany, H. Qi, “An energy-efficient QoS-aware
media access control protocol for wireless sensor networks”, in
IEEE International Conference on Mobile Adhoc and Sensor
Systems,2005, pp. 189-191.
[8] IEEE Standard for Wireless LAN Medium Access Control
(MAC) and Physical Layer (PHY) Specifications, IEEE 802.11,
1999.
[9] H. Zhai, J. Wang, and Y. Fang, "Distributed packet scheduling
for multi-hop flows in ad-hoc networks,” in Proc. of WCNC,
Atlanta, GA, USA, 2004. pp. 1081-1086.
[10] K. Sundaresan,V. Anantharaman, H.Y. Hsieh, R.Sivakumar
"ATP: A reliable transport protocol for ad-hoc networks," in
proc. of MOBIHOC, NY, USA, 2003.pp. 64-75.
[11] S.B. Lee, J. Cho, and A. T. Campbell. (2003). A hotspot
mitigation protocol for ad hoc networks, Ad Hoc Networks,
vol. 1, pp. 87-106.
[12] B. Ngo, S. Gordon,“Avoiding Bottlenecks Due To Traffic
Aggregation At Relay Nodes In Multi-Hop Wireless
Networks”,in Proc. Of IEEE Asia Pacific Conference on
Communication, Perth, WA. 2005. pp. 769-773.
[13] R.K Tripathi, R. Shakya, N.K Verma, YN Singh, “Localized
detection of bottleneck nodes and quantification of criticality in
a wireless sensor networks” IEEE SIBIRCON, Listvyanka.
2010.pp. 326-328.
[14] L. Tao, F.Yu.(2010, August) ECODA: Enhanced Congestion
Detection and Avoidance for Multiple Class of Traffic in
Sensor Networks. IEEE transaction on consumer electronics.
Vol. 56(3).pp. 1387-1394.
[15] Davide Astuti (2003, August). Packet Handling, Seminar on
Transport of Multimedia Streams in Wireless Internet, Helsinki,
Finland.
[16] Nidal Nasser, Lutful Karim, and Tarik Taleb.(April, 2013).
Dynamic Multilevel Priority Packet Scheduling Scheme for
Wireless Sensor Network. IEEE transactions on Wireless
Communications, vol.12 (4).pp. 1448-1459.
[17] J. S. Hartmut Ritter, Rolf Winter, “A Partition Detection
System for Mobile Ad-hoc networks, “in Sensor and Ad Hoc
Communications and Networks (IEEE-SECON), 2004. pp.
489–497.
[18] M.H. D.S.R. Milenko Jorgic, Ivan Stojmenovic, “Localized
Algorithms for Detection of Critical Nodes and Links for
Connectivity in Ad Hoc Networks,” in 20th International
Conference on Advanced Information Networking and
Applications Volume 2 (AINA’06).USA. 2006. pp. 336–340.
[19] Network Simulator-NS2. Available at: www.isi.edu/nsnam/ns/.
[20] Generating node-movement and traffic-connection files for
large wireless scenario.
Available:http://www.isi.edu/nsnam/ns/tutorial/nsscript7.html.
[21] Chuck Semeria, “Supporting Differentiated Service Classes:
Queue Scheduling Disciplines”,2001, Juniper Networks.
[22] S. Bansode and S. Sambare, “Performance Evaluation of
Dynamic Multilevel Priority (DMP) Packet Scheduling Method
for Wireless Sensor Networks (WSNs)” in Proc. IEEE
International
Conference
of
Pervasive
Computing
(ICPC),India. 2015,pp 1-6.
Page 42
[23] M.H. Taha, N.E.D.M. Khalifa, H.N. Elmahdy and I.A
Saroit(2013), “Energy Based Scheduling Scheme for Wireless
Sensor Networks” CiiT International Journal of Wireless
Communication, vol.4(16),pp. 973-978.
[24] J. Liu, N. Gu and S. He, “An Energy-Aware Coverage Based
Node Scheduling Scheme for Wireless Sensor Networks “ in
International Conference for Young Computer Scientists,
ICYCS, 2008,pp. 462-468.
[25] E. Srie Vidhya Janani and P. Ganesh Kumar (2015). Energy
Efficient Cluster Based Scheduling Scheme for Wireless Sensor
Networks. The Scientic World Journal. Volume 2015. Article
ID 185198.
ISSN – 1999-4974