Please write detailed REVIEW on Following Topic. It required only one page.
koti
fair-queue-MANET1.pdf
For R eview
O nly
Energy Efficient MAC Protocol with Fair-Scheduling
Technique in Multi-hop MANETs
Journal: KSII Transactions on Internet and Information Systems
Manuscript ID: TIIS-RP-2014-Jan-0089
Manuscript Type: Regular Paper
Date Submitted by the Author: 20-Jan-2014
Complete List of Authors: Sivanesan, P. Thangavel, S.
Keywords of your Paper: Computer Networks, Computer Security, Cooporative Communications/Networking
http://mc.manuscriptcentral.com/tiisjournal
KSII Transactions on Internet and Information Systems
For R eview
O nly
Energy Efficient MAC Protocol with Fair- Scheduling Technique in Multi-hop
MANETs
1 P.Sivanesan and
2 S.Thangavel
1 Associate Professor
Department of Information Technology,
M.A.M College of Engineering, Tiruchirappalli.
2 Professor, Department of EEE
K.S.Rangasamy College of Technology, Tiruchengode
Abstract
In Mobile Ad-hoc Networks (MANETs), achieving fairness in bandwidth allocation and
increasing channel utilization are the important design goals of scheduling. However, these
two goals contradict with each other. In this paper, a fair-scheduling technique for inelastic
traffic flows in MANET is proposed. The network traffic is differentiated into two categories
as elastic and inelastic flows. In this technique, data packets of inelastic flows are prioritized over data packets of elastic flows. Utility function is estimated considering channel
utilization, channel state information along with delay of data packets. When more than a
data packet of inelastic flows compete in scheduling, packet with high(upper bound) on
delay field is prioritized and scheduled. The proposed technique is validated through
simulation results. It proves that our proposed technique offers fairness in scheduling
network traffic.
Key Words: Channel utilization, Fair-scheduling, MAC, MANET, Prioritized.
1. Introduction
1.1 Mobile Ad-hoc Networks (MANETs)
A self-governing system with a set of mobile nodes that are permitted to move randomly in
the network is defined as a Mobile Ad-hoc Network (MANET). Since, the network lacks
Page 1 of 13
http://mc.manuscriptcentral.com/tiisjournal
KSII Transactions on Internet and Information Systems
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
For R eview
O nly
centralized database or communication infrastructure, every node forwards the data packet to
the destination through a multihop radio link [1] [2]. The ability of MANET to operate
without the need of any fixed infrastructure leads to many appealing applications such as
military, disaster recovery, trucks, airplanes, ships etc [2].
1.2 Scheduling in MANET
The task of scheduling is to decide on the processing of next appropriate packet in the queue
such that it enhances the end-to-end performance even when the network traffic load is high
[3]. In Time Division Multiple Access (TDMA) method, time slots are used to transmit and
receive data. Here, the basic unit of scheduling is a frame [1]. With reference to the Open Systems Interconnection (OSI) model, the scheduler is situated connecting the routing agent
and the Media Access Control (MAC) layer. Packet scheduling determines which flow must
be serviced among the group of waiting flows. Packet scheduling plays a vital role in
offering Quality of Service (QoS) guarantees. However, it is challenged by the mobile and
dynamic nature of the nodes [4].
In MANET, an ideal scheduling algorithm should be accustomed with dynamic changing
of topology and limited bandwidth [4]. Since, wireless networks become predominant it is
anticipated to sustain various services such as best effort and real-time traffic. These
networks have to meet with QoS requirements like minimum bandwidth and maximum delay
limitations while still maintaining network queues stable and enhancing network throughput
[5].
In mobile wireless ad hoc networks, scheduling medium access is a demanding process as
a result of node’s mobility, limited availability and constrained bandwidth. Using TDMA
networks, each node is allocated with a fixed length of time slot to transmit data. However,
this method is applicable only when the network size is small and nodes are aware of
network connectivity [6].
1.3 Requirements for Fair Scheduling Algorithm
In Ad-hoc network, an ideal fair scheduling algorithm for network layer requires the following characteristics [7],
• Priority queues have to be established to handle newly entering flows as per their priorities
• Fairly allocate network layer resources among various flows in order to meet the fairness in allocation
• An ideal fair scheduling algorithm should be well-suited to all routing algorithms
• It must enhance the overall network throughput
• The scheduling algorithm should remain stable in high load situation and it must be adaptable to total network delay.
1.4 Issues of Scheduling in MANET
Achieving fair bandwidth allocation and increasing channel utilization are the important
design goals of packet scheduling in Mobile Ad-hoc Networks (MANETs). Nonetheless,
these two goals are contradict each other. Some of the challenges of scheduling in MANET
are listed below,
Page 2 of 13
http://mc.manuscriptcentral.com/tiisjournal
KSII Transactions on Internet and Information Systems
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
For R eview
O nly
• Scheduling becomes more complex task when it is performed along with power control issues, for nodes in Ad-hoc network can communicate with several other
nodes simultaneously [9].
• Accomplishing bandwidth maximization and fairness at hand is a difficult problem in MANET as it uses shared-medium. Apart from this, issues such as distributed
nature of packet scheduling, spatial channel reuse and location-dependent contention
among flows complicate scheduling in MANET [8].
• Typically, during scheduling, in order to achieve high throughput some data flows are delayed, which causes unfairness in flows [8].
• Stringent battery power, hidden terminal problem and error susceptible communication channel are the other different factors that affect scheduling and end-
to-end packet delivery in MANET [10].
1.5 Problem Identification
In our previous work [17], we have proposed an energy efficient MAC protocol in MANET
based on channel utilization and queue size. In this protocol, the source node that desires to
transmit the data packet to its destination node, appends its queue state and channel
utilization with request to send (RTS) frame and transmits it to the destination utilizing the
utility function. After the destination verifies the RTS frame for error, its sends the clear to
send (CTS) frame along with queue state and channel utilization information to source node
using utility function. Upon receiving CTS frame, the source transmits the data packet to the
destination node.
2. Related Work
Sang-Chul Kim [1] has introduced an energy efficient scheduling algorithm for cluster based
mobile wireless networks. Their algorithm has utilized the adaptive –tier low transmission power schedule strategy in TDMA based AdHoc MAC protocol. In their algorithm,
transmission power is calculated by estimating the distance between Cluster Head (CH) and
non-CH (NCH) nodes and as a result path-loss is gradually reduced. Further, they have
proposed a scheme for interference avoidance. Using this, when neighbor clusters transmit
packets, the total energy dissipation is minimized and it increases the utilization of time slot
in each ad hoc node.
Juan Jos´e Jaramillo and R. Srikant [5] have presented an optimal scheduling scheme for
fair resource allocation in AdHoc network with elastic and inelastic traffic. Their model
integrates the QoS requirements of packets with deadlines in the optimization framework.
Thus, their technique provides a solution for both congestion control and scheduling scheme
that fairly allocates resources to satisfy requirements of both elastic and inelastic flows. Their
algorithm has been derived through a dual decomposition approach.
A Fair Data Flows Scheduling Schema for Multihop Wireless Ad Hoc Networks is
proposed in [7]. Their main objective is to allocate bandwidth fairly among different
contention traffic flows. They have used a novel History Based Priority Queuing (HBPQ)
algorithm for scheduling different traffic flows. Their HBPQ makes use of satisfaction
function to calculate user’s satisfaction and then attempts to bring close to the satisfaction of
Page 3 of 13
http://mc.manuscriptcentral.com/tiisjournal
KSII Transactions on Internet and Information Systems
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
For R eview
O nly
users. According to this technique, every data packet is allocated with the services based on
its experienced traffic load in traveled path.
Joint per-flow scheduling and routing technique in wireless multihop networks is
proposed in [11] by Dimitrios J. Vergados et al. Their scheduling and routing algorithm
routes the flows in a manner that stays away from congested area with limited availability.
Instead of assigning slots to links or nodes, their scheduling mechanism assigns to flows. It
fairly assigns slots and assures the fairness in allocation does not lead to underutilization.
Dang-Quang Bui and Won-Joo Hwang [12] have put forwarded a proportionally Quasi-
Fair Scheduling Optimization framework in Wireless Ad Hoc Networks. They have
developed their optimization framework with the objective of assuring fairness of the
cumulative data rates. Their scheduling algorithm is asymptotically stable. Making use of
stochastic process analysis, they have proved that the cumulative rates generated by the
framework converge to the unique limit point of an ordinary differential equation. Their
scheme can also be enhanced for throughput maximization and max-min fairness schemes.
Vijay S Rao et al. [13] have proposed a maximizing the fair allocation of opportunistic
spectrum for Cognitive Radios (CRs) Adhoc networks. Initially, they have addressed the
issue of fairness in a CR ad hoc network and then designed a novel distributed heuristics to
allocate spectrum fairly. They have framed the problem of fair allocation of channels in a CRAN in a time slotted framework. They have proposed their scheme with the intention of
all transmitter-receiver pairs should approximately be allocated with the same percentage of
access to the medium over sufficient time.
3. Proposed Solution
3.1 Overview
In this paper a Fair-Scheduling Technique for Inelastic Traffic Flows in MANET is proposed. The data traffic is differentiated into two categories as elastic and inelastic traffic
considering their application requirements. The proposed technique prioritizes data packets
of inelastic flow over data packets of elastic flow. This is achieved as inelastic flow requires
maximum delay requirements. Utility function of nodes is measured considering channel
utilization, channel state information and packet delay value. When data packets arrives the
scheduler, it initially differentiates them according to the type of flow it belongs. Data
packets of inelastic flows are sorted as per their delay field in utility function. The data
packet with high delay value is prioritized and scheduled first.
3.2 Estimation of Channel Utilization
Let dist (t) be the distributed inter-frame spacing (DIFS). Let s (t) be the short inter-frame
spacing SIFS. Let r (t) represents the Request to send (RTS) control packet. Let c (t)
represents the Clear to send (CTS) control packet. Let d (t) be the data packet. Let a (t) be the
acknowledgement frame (ACK). Let DDIFS, DSIFS, DRTS, DCTS, Dd, DACK be the delay
components of DIFS, SIFS, RTS, CTS, data and ACK packets respectively.
The channel utilization of a network per second is computed using the following
components [14]
• The time taken for transmission and management of the data packets and control frames
Page 4 of 13
http://mc.manuscriptcentral.com/tiisjournal
KSII Transactions on Internet and Information Systems
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
For R eview
O nly
• The total number of delay components such as Distributed inter-frame spacing (DIFS) and short inter-frame spacing (SIFS).
The estimation of above components and estimation of channel utilization based on these
components are illustrated below.
The channel busy time (Tchb) for a data frame (d) is given using Eq: (1)
Tchb (d) = DDIFS + Dd (z,σ ) (1) Where z = size of data frame (in bytes)
σ = rate at which the data is transmitted. DDIFS = delay component
Tchb of the RTS frames is given using Eq: (2)
Tchb (RTS) = DRTS (2)
Tchb of the CTS frames is given using Eq: (3)
Tchb (CTS) = DSIFS + DCTS (3)
Tchb of the ACK frames is given using Eq: (4)
Tchb (ACK) = DSIFS + DACK (4)
If RTS, CTS, ACK and data packets are encountered during the interval t, the total Tchb
is given using the following equation(5).
Tchb (t) = (r (t) * Tchb (RTS)) + (c (t) * Tchb (CTS)) + (a (t) * Tchb (ACK)) + (d (t) * Tchb
(d)) (5)
Thus, the Channel utilization at time t, is given using equation (6).
CU (t) = 100*
10
)( 6
tT chb
(6)
3.3 Estimation of Channel State Information
The channel state information refers to the physical layer information that includes channel
fading, multipath propagation, reflection, scattering and other climatic effects on the channel.
It is estimated based on the signal strength and signal to noise ratio (SNR) at the receiver.
The estimation of signal strength using Friis equation is given by Eq: (7) [15]
τσ
σβα
*)**4(
***** Pr
2
2
d
hhP x rxtxtx= (7)
Where Ptx = transmitted power , α = transmitter gain, β = receiver gain, htx = transmitter height, hrx = receiver height,
σ = wavelength , d = distance among the transmitter and receiver, τ = system loss .Based on the estimation of signal strength, SNR is computed using equation (8)
SNR = log10(Ptx) – log10(Prx) dB (8)
3.4 Estimation of Flow Delay
The proposed technique estimates the packet delay as a difference between the times. It is
transmitted from the source to the time it reaches the destination. Let tPi be the time the data packet Pi is transmitted at the source and rPi be the time the data packet Pi is received at the
destination. Now, the packet delay of data packet Pi is computed as,
iii rPtPP −=
(9)
Page 5 of 13
http://mc.manuscriptcentral.com/tiisjournal
KSII Transactions on Internet and Information Systems
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
For R eview
O nly
Thus, flow’s delay ( fD
) is the average of its packet’s delay. Accordingly, the network’s
delay (nD ) is the sum of fDof all active flows in the network.
It can be given as,
nD= ∑ =
n
i
i Df
1 where, ActiveFlowf
i =
(10)
3.5 Estimation of Utility Function
In this paper, we enhance the utility function of our previous work [17] to prioritize data
flows while scheduling. Considering channel state information, cost factor and flow delay,
the transmitter node estimates the utility function. Once the utility function is estimated, by
looking at the delay field of a flow, the scheduler prioritizes the inelastic flows, which will
be discussed in later section. Thus, the utility function helps in prioritizing inelastic flows
and improves network fairness [16].
According to Shannon’s capacity, the utility function for an active link i is defined as
UFi = ln (1 + ) i
SNR - CFi Ptxi +
Df i i = 1,2,…,n. (11)
where Ptxi = transmission power of the node
Ptx-i = transmission power of all links other than i
∆
=[Ptx1, …., Ptxi-1, Ptxi+1, …. Ptxn)
i SNR
= Signal to noise ratio (SNR) received at the receiver node.
CFi = cost factor representing the specific amount of power consumption.
Df i = flow’s delay
3.6 Network Model
Consider the network as a directed graph ( )LNG ,
where N represents the set of nodes and
L denotes the set of directional links. Let n1, n2 … N be the set of mobile nodes and L1,
L2… L be the set of links that connect the network. Network traffic is considered as the mixture of both elastic and inelastic traffic flows. We are supposing that traffic flows are
described and differentiated as elastic and inelastic by the application corresponding to its
requirements.
3.7 Scheduling Scheme
In this technique, data flows transmitted over the communication channel of MANET are
divided into two categories as elastic and inelastic traffic flows. An inelastic traffic denotes
important applications such as real-time traffic, delay adaptive traffic and rate adaptive
traffic. Therefore, it has stringent maximum per flow packet delay requirements. On the
other hand, an elastic flow does not require any delay requirements. Since, they are elastic
flows; they are able to reduce their transmission rate and they can tolerate packet delays and
packet losses elegantly.
Page 6 of 13
http://mc.manuscriptcentral.com/tiisjournal
KSII Transactions on Internet and Information Systems
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
For R eview
O nly
During the phase of scheduling, the time is split into slots. A set of ‘t’ successive time
slots constitutes a frame. The scheduling technique presupposes that data packets arrive only
at the beginning of a frame and flows that contain inelastic packets have‘t’ time slots as their
deadline.
Take into account that in Fa
as the number of inelastic packets enters at each link and it
can be represented as,
( ) iinin LFaFa =
where LL
i ∈
(12)
Here, in Fa
is the random variable with mean λ and variance 2σ .
In a given frame, assume e Fa
as the number of elastic packets entered at each link. We
presume that the channel state is constant for a given frame, which is independent of
different frames and arrival of data packets. The vector ‘ch’ represents the total number of
packets that can be successfully transmitted on link Li in a given frame. It is symbolized as,
( ) LLchch
iLi ∈=
(13)
At time T, for a link LL
i ∈
, consider shas a possible schedule that denotes,
( )
TLeTLin ii shshsh
),(),( ,=
(14)
Where, TLeTLin ii shsh
),(),( ,
denote the possible number of inelastic and elastic packets that
can be scheduled at link Li L∈
in timeslot T. Therefore, if TLeTLin ii shsh
),(),( +
> 0 then the
link Li is scheduled to transmit data in timeslot (T) and also when TLeTLin shsh
),1(),1( +
>0and
TLeTLin shsh
),2(),2( +
> 0, then the links L1 and L2 can be scheduled and transmitted at the
same time without intruding each other.
3.8 Fair-Scheduling Technique for Inelastic Traffic Flows in MANET
In this fair scheduling technique, inelastic flows are prioritized over elastic flows. As we
mentioned above, the flows are differentiated into elastic and inelastic flows as per the
application requirement. Once a packet of a flow is transmitted to the destination, the utility
function is computed. While packets are arriving, the scheduler looks for the type of flow
(elastic (or) inelastic). If the packet belongs to elastic flow, the scheduler allocates time slots
according to its feasibility. On the other hand, if the packet belongs to inelastic flow then immediately the scheduler checks the delay field in utility function. The scheduler gives high
priority to them.
When more than one packet belongs to inelastic flow, the scheduler sort the packets
based on their delay field. Packets with high delay value are placed in front line of the queue.
Since, inelastic flows necessitate maximum delay requirements they are scheduled first by
the scheduler when comparing elastic flows. When more than a packet belongs to inelastic
flow, then the packet with high delay field is scheduled first.
Assume that at time ‘t’, dP1 of F1, dP2 of F2 and dP4 of F3 arrive the scheduling queue.
Once the data packets are reached the queue, the scheduler overseers the data packets in
order to discover the type of flow.
Page 7 of 13
http://mc.manuscriptcentral.com/tiisjournal
KSII Transactions on Internet and Information Systems
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
For R eview
O nly
Figure 1. Connected Network.
From Table-1, we could understand that dP1 belongs to elastic flow and dP2 and dP4
correspond to inelastic flows. Therefore, the scheduler prioritizes dP2 and dP4 over dP1.
Since, dP2 and dP4 belong to inelastic flows; the scheduler checks for the delay field of dP2
and dP4 from their utility functions. We also suppose that dP4 has high delay value than
dP2. Therefore, dP4 is scheduled first, then dP2 is scheduled and finally dP1 is scheduled
finally. The described process is illustrated in figure 2.
Table 1. Elastic and Inelastic flows
Data packets Flow id Type of flow
dP1 F1 Elastic
dP2, dP3 F2 Inelastic
dP4, dP5, dP6 F3 Inelastic
dP7 F4 Inelastic
dP8 F5 Elastic
dP9 F6 Elastic
dP10, dP11 F7 Inelastic
dP12 F8 Elastic
Figure 2. Prioritized flow scheduling
The proposed fair-scheduling algorithm for Inelastic Traffic Flows in MANET is
described below in the following algorithm.
1. Let Fel denotes elastic flow and Fin represents inelastic flow.
Page 8 of 13
http://mc.manuscriptcentral.com/tiisjournal
KSII Transactions on Internet and Information Systems
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
For R eview
O nly
2. Assume N as the total number of nodes in the network.
3. Let dPi be the ith data packet i = 1, 2 … n
4. Nodes n1, n2…N are initialized in the network
5. Data flows are described as Fel and Fin considering application requirements
6. Utility function of nodes are calculated as per equation(11)
7. At the time ‘t’, dP’s of different flows are transmitted over the communication
channel
8. Transmitted dP’s reach the scheduling queue
9. The scheduler differentiates dP’s according to their type of flow they belongs
10. If (flow (dPi) = Fin) then
The dPi is prioritized over other data packets
11. If (dPi = 1) then
11.1 The scheduler schedules
dPi with requiring timeslots
12. Else if (dPi > 1) then
13. The scheduler looks utility function of nodes and discovers delay value.
14. dP’s are sorted according to their delay value
15. dPi that has high delay value is allocated first
16. Else if (flow (dPi) = Fel) then
16.1 The scheduler looks for other dP’s that correspond to inelastic flow
17. If (There is no inelastic flow data packets) then
18. dP of elastic flow is scheduled 19. Else
20. dP is scheduled with minimum requirements
21. End if
4. Simulation results
4.1 Simulation Model and Parameters
The Network Simulator (NS2) [18], is used to simulate the proposed architecture. In the
simulation, 100 mobile nodes move in a 1000 meter x 1000 meter region for 100 seconds of simulation time. All nodes have the same transmission range of 250 meters. The simulated
traffic is Constant Bit Rate (CBR).
Table 2. Simulation parameter settings
No. of Nodes 100
Area Size 1000 X 1000
Mac AEEMAC
Transmission Range 250m
Simulation Time 100 sec
Traffic Source CBR
Packet Size 512
Initial Energy 22.1J
Transmission Power 0.660
Receiving Power 0.395
Rate 100,200,300,400 and 500Kb
Routing Protocol AODV
Page 9 of 13
http://mc.manuscriptcentral.com/tiisjournal
KSII Transactions on Internet and Information Systems
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
For R eview
O nly
4.2 Performance Metrics
The proposed Energy Efficient MAC protocol With Fair-Scheduling Technique
(EEMACFS) is compared with the OSFRA technique [ ]. The performance is evaluated
mainly, according to the following metrics.
Packet Delivery Ratio: It is the ratio between the number of packets received and the
number of packets sent.
Packet Drop: It refers the average number of packets dropped during the transmission
Throughput: It is the total number of packets received by the receiver.
Delay: It is the amount of time taken by the nodes to transmit the packets to the receiver. In this experiment, we vary the transmission rate as 100,200,300,400 and 500Kb based on
data rate.
Rate Vs Received Bandwidth
0
2
4
6
8
100 200 300 400 500
Rate(Kb)
M b /s EEMACFS
OSFRA
Figure 3. Rate Vs Received Bandwidth
Rate Vs Delay
0
10
20
30
100 200 300 400 500
Rate(Kb)
D e la y (S e c )
EEMACFS
OSFRA
Figure 4. Rate Vs Delay
Rate Vs DeliveryRatio
0
0.5
1
1.5
100 200 300 400 500
Rate(Kb)
D e li v e ry R a ti o
EEMACFS
OSFRA
Figure 5. Rate Vs Delivery Ratio
Page 10 of 13
http://mc.manuscriptcentral.com/tiisjournal
KSII Transactions on Internet and Information Systems
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
For R eview
O nly
Figure 3 shows the received bandwidth of EEMACFS and OSFRA techniques for
different transmission rate scenario. We can conclude that the received bandwidth of our
proposed EEMACFS approach has 87% of higher than OSFRA approach. Figure 4 shows
the delay of EEMACFS and OSFRA techniques for different transmission rate scenario. We
can conclude that the received delay of our proposed EEMACFS approach has 49% of less
than OSFRA approach. Figure 5 shows the delivery ratio of EEMACFS and OSFRA
techniques for different transmission rate scenario. We can conclude that the delivery ratio of
our proposed EEMACFS approach has 83% of higher than OSFRA approach. Figure 6
shows the packet drop of EEMACFS and OSFRA techniques for different transmission rate
scenario. We can conclude that the packet drop of our proposed EEMACFS approach has
40% of less than OSFRA approach.
Rate Vs Drop
0
5000
10000
15000
100 200 300 400 500
Rate(Kb)
P k ts
EEMACFS
OSFRA
Figure 6. Rate Vs Drop
Rate Vs ResidualEnergy
0
5
10
15
100 200 300 400 500
Rate(Kb)
E n e rg y (J )
EEMACFS
OSFRA
Figure 7. Rate Vs Residual Energy
Rate Vs Throughput
0
5000
10000
15000
20000
100 200 300 400 500
Rate(Kb)
T h ro u g h p u t
EEMACFS
OSFRA
Figure 8. Rate Vs Throughput
Figure 7 shows the residual energy of EEMACFS and OSFRA techniques for different
transmission rate scenario. We can conclude that the residual energy of our proposed
EEMACFS approach has 16% of higher than OSFRA approach. Figure 8 shows the
Page 11 of 13
http://mc.manuscriptcentral.com/tiisjournal
KSII Transactions on Internet and Information Systems
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
For R eview
O nly
throughput of EEMACFS and OSFRA techniques for different transmission rate scenario.
We can conclude that the throughput of our proposed EEMACFS approach has 87% of
higher than OSFRA approach.
5 Conclusion
In this paper, we have put forwarded a Fair-Scheduling Technique for Inelastic Traffic Flows
in MANET. The network traffic is differentiated into two categories as elastic and inelastic
flows. In this technique, data packets of inelastic flows are prioritized over data packets of
elastic flows. Utility function is estimated considering channel utilization, channel state information along with delay of data packets. When data packets arrives the scheduler, it
initially differentiates them according to the type of flow it belongs. Data packets of inelastic
flows are sorted as per their delay field in utility function. The data packet with high delay
value is prioritized and scheduled first. The proposed technique is validated through
simulation results. It proves that our proposed technique offers fairness in scheduling
network traffic.
References
[1] Sang-Chul Kim, “An Energy Efficient Scheduling Algorithm for Cluster- Based Mobile
Wireless Networks”, IEEE Fourth International Conference on Ubiquitous and Future
Networks (ICUFN), 2012
[2] Nur Idawati Md Enzai, Farhat Anwar and Omer Mahmoud, “Evaluation Study of QoS-
Enabled AODV”, Proceedings of the International Conference on Computer and Communication Engineering, 2008.
[3] Kanika Garg and RishiPal Singh, “Scheduling Algorithms in Mobile Ad Hoc
Networks”, The International Journal of Computer Science & Applications (TIJCSA),
Volume 1, No 5, 2012.
[4] Nur Idawati Md Enzai, Siti Sara Rais and Rosheila Darus, “An Overview of Scheduling
Algorithms in Mobile Ad-Hoc Networks”, International Conference on Computer
Applications and Industrial Electronics (ICCAIE 2010), 2010.
[5] Juan Jos´e Jaramillo and R. Srikant, “Optimal Scheduling for Fair Resource Allocation
in Ad Hoc Networks with Elastic and Inelastic Traffic”, IEEE INFOCOM, 2010.
[6] Yiming Liu, Victor O. K. Li, Ka-Cheong Leung, and Lin Zhang, “Topology-
Transparent Distributed Multicast and Broadcast Scheduling in Mobile Ad Hoc
Networks”, Proceedings of IEEE VTC 2012.
[7] Hamidreza Salarian and Pejman Khadivi, “Fair Data Flows Scheduling Schema for
Multihop Wireless Ad Hoc Networks”, 16th Asia-Pacific Conference on
Communications (APCC), 2010.
[8] Tao Ma, Xinming Zhang and Guoliang Chen, “A Maximal Clique Based Packet
Scheduling Algorithm in Mobile Ad Hoc Networks,” IEEE ICN'04, pp-690-695, Feb,
2004.
[9] Sneha S. Tirth, S.M.Jagade & Shilpa Sapatnekar, “Scheduling Techniques for MANET
with CDMA Protocol”, International Conference On Engineering Innovation and Technology, ISBN: 978-93-81693-77-3, Nagpur, 1st July, 2012.
[10] D. Anitha, and Dr.M.Punithavalli, “Secure MANET Communication Using Cluster
Based Multicast with Fair Key and Resource Scheduler” IJCSN International Journal
of Computer Science and Network, Vol-2, Issue-1, 2013.
Page 12 of 13
http://mc.manuscriptcentral.com/tiisjournal
KSII Transactions on Internet and Information Systems
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
For R eview
O nly
[11] Dimitrios J. Vergados, Aggeliki Sgora, Dimitrios D. Vergados and Periklis
Chatzimisios, “Joint Per-Flow Scheduling and Routing in Wireless Multihop
Networks”, IEEE Symposium on Computers and Communications (ISCC), 2011.
[12] Dang-Quang Bui and Won-Joo Hwang, “Proportionally Quasi-Fair Scheduling
Optimization in Wireless Ad Hoc Networks”, 26th IEEE International Conference on
Advanced Information Networking and Applications, 2012.
[13] Vijay S Rao, R Venkatesha Prasad, R Muralishankar, and I.G.M.M. Niemegeer,
“Maximizing the Fair Allocation of Opportunistic Spectrum for CR Ad hoc Networks”,
IEEE Global Telecommunications Conference (GLOBECOM 2011), 2011.
[14] Amit P. Jardosh Krishna N. Ramachandran Kevin C. Almeroth Elizabeth M. Belding-
Royer, “Understanding Congestion in IEEE 802.11b Wireless Networks”, Proceeding
of the 5th ACM SIGCOMM conference on Internet Measurement, 2005.
[15] Jagadeesh Kumar, S.J.K. and T. Purusothaman, “A Cross Layer Frame Work for
Enhanced Quality of Service Provisioning in Worldwide Interoperability for Microwave
Access Networks”, Journal of Computer Science, pp 420-424, 2012.
[16] Fan Wang, Ossama Younis and Marwan Krunz, “Throughput-oriented MAC for mobile
ad hoc networks: A game-theoretic approach”, IEEE, Volume 7, Issue 1, Pages 98–117,
January 2009.
[17] P.Sivanesan and Dr.S.Thangavel, “Achieving Energy Efficient MAC protocol for
MANETs by Channel Utilization and Queue Size”, International Review on
Computers and Software (IRECOS), April 2013, Vol. 8. n. 4, pp. 996-1003.
[18] Network Simulator: http:///www.isi.edu/nsnam/ns.
Page 13 of 13
http://mc.manuscriptcentral.com/tiisjournal
KSII Transactions on Internet and Information Systems
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
