Cross Layer Based Hybrid Fuzzy Ad-Hoc Rate Based Congestion Control (CLHCC) Approach for VoMAN to Improve Quality of VoIP Flows

Cross Layer based Hybrid fuzzy ad-hoc rate
based Congestion Control (CLHCC) approach
for VoMAN to improve quality of VoIP flows
V.Savithri1
,1 Assistant Professor, Part-Time Ph.D. Scholar, Bharathiar University, Coimbatore, India.
Dr.A.Marimuthu2
, Associate Professor& Head, PG and Research Department of Computer Science,
Government Arts College(Autonomous),Coimbatore, India.
Abstract - Mobile environment pretense a number of novel
theoretical and optimization issues such as position, operation
and following in that a lot of requests rely on them for
desirable information. The precedent works are sprinkled
across the entire network layer: from the medium of physical
to link layer to routing and then application layer. In this
invention, we present outline solutions in Medium Access
Control (MAC), data distribution, coverage resolve issues
under mobile ad-hoc network environment based on
congestion control technique using Transmission Control
Protocol (TCP). In mobile ad-hoc network issues can arise
such as link disconnections, channel contention and recurrent
path loss. To resolve this issue, we propose a Cross Layer
based Hybrid fuzzy ad-hoc rate based Congestion Control
(CLHCC) approach to maximize network performance. Based
on the destination report it regulates the speed of data flow to
control data loss by monitoring the present network status
and transmits this report to the source as advice. The source
adjusts the sending flow rate as per the advice. This is
monitored by channel usage, ultimate delay, short term
throughput.
Index Terms— MAC, MPSD, RTT, SIFS TCP,
I. INTRODUCTION
Mobile Ad-hoc Network (MANET) is a
transient dynamic network formed by set of mobile nodes.
Routing in MANTE is a complex and challenging task
because of its dynamic nature, link stability and
infrastructure less concept. A packet might traverse many
intermediate nodes supporting dynamic link to reach the
desired destination. Routing algorithm should generate
feasible route by collecting routing status and route packets
over the optimal route to support Quality of Service (QoS).
Voice over MANET is becoming important day
by day as users demand to use real time applications. As
Quality of Service (QoS) issues in infrastructure based
networks still remain unsolved, it is a challenging task in
MANET that needs to be solved for real time applications
[12]. Increase in real time traffic increases network
congestion. Congestion control policies are classified into
three types namely window based, rate based and hybrid.
Window based congestion control policy adjusts the
congestion window as per the changing network status.
Rate based congestion control policy increases or decreases
the data rate of the sender as per the current status of the
network [8]. Hybrid approach combines both the above
discussed policies to control congestion.
International Journal of Computer Science and Information Security (IJCSIS),
Vol. 16, No. 4, April 2018
289 https://sites.google.com/site/ijcsis/
ISSN 1947-5500

This paper is organized as follows: Section II describes the
literature survey and Section III presents the methodology
of Fuzzy based Congestion Control approach using voice
over TCP. Section IV presents the result analysis of the
new approach. Section V summarizes the most important
simulation results and their interpretation. Finally, Section
VI concludes this paper.
II. LITERATURE SURVEY
Security in mobile ad hoc network is hard to
accomplish due to vibrantly changing and fully
decentralized topology as well as the vulnerabilities and
limitations of wireless data transmissions[15].A lot of
research has since focused on mechanisms to improve TCP
performance in cellular wireless systems[1]. In MANET,
due to dynamic nature of the network and variable number
hops between source and destination, the fairness and
efficiency get degraded. It is also seen that the existing
protocols are not good enough in setting its parameters like
congestion window, congestion window limit, round trip
time and the retransmission timeout timers [11]. The
routing overhead of such an algorithm increases with the
square of the number of mobile nodes in a MANET[2]. A
hybrid routing algorithm that combines the merits of
existing protocols that can be used to address this issue of
growth in network size and load balancing whose behavior
can be modified according to the size of network [9]. The
effect of high bit error rate and route re-computation on the
performance of TCP in mobile ad hoc network is analyzed
[4]. In particular, it allows the routing protocol to operate
more efficiently by reducing the control traffic in the
network and simplifying the data routing [7]. In order to
avoid congestion in the network, it is required to use an
efficient congestion control algorithm for successful
transmission of data throughout the network [13].
Numerous examinations have been done to enhance the
execution of TCP at network layer by enhancing the
routing strategy. M-ADTCP is another method that
presents a Modified AD-hoc Transmission Control
Protocol where the receiver detects the probable current
network status and transmits that information to the sender
as feedback. The sender behaviour is altered appropriately
[6]. In this way, there is an extension to enhance the
execution of TCP at transport layer [14]. TCP supplies end-
to-end reliable delivery of data between an application
process on one computer and the process running on
another computer, by adding services on top of IP [5]. In
order to use these limited resources efficiently, an
intelligent routing strategy is required which should also be
adaptable to the changing conditions of the network, like,
size of the network, traffic density and network
partitioning[10].
III. A.CLHCC METHODOLOGY
Application Layer
VoIP based Communication
Transport Layer
Congestion detection and Controller
System
Network Layer
Computation of Middle Packet Setback
Distinction, Short Term Throughput,
Delay
Physical and MAC Layer
Computation of Channel business ratio
Figure 1. Cross Layer based VoMAN Architecture
We consider a mobile ad-hoc network with n
nodes S (s1; s2; . . .sn;). These n nodes are randomly
disseminated in a field. At random time periods, each node
Vol. 16, No. 4, April 2018
ISSN 1947-5500

can send event values ( VE ) from the field at its location
and sends them VE to the destination. Nodes send a
network wide broadcast message at every particular
interval, to all nearby nodes in the network. Nodes built
with a global positioning system (GPS) recipient at an
exactly known position which is used to receive collected
information from nearby nodes. The nodes are installed
randomly and they may organize by them self as hop based
jointly to complete a transmission sensed the environment.
Detection of nearby nodes is accomplished by a periodic
HELLO messages sharing among the nodes, produced at
fixed interval. When a node receives HELLO message it
updates the neighbor table NT.
Figure 2. Transmission Area
If node does not receive HELLO message within
the fixed time interval at each sequence purge it from NT.
By this way update nearby routing node freshly. Mobile
devices are deployed in the network region randomly and
moves anywhere in the network using the random way
point mobility model in non uniform manner. Source node
initiates the data transmission by sending the first control
packet to the destination.
In network layer, device looks for the routing table
(RT) entry to forward the packet to the destination. If there
is no route found in RT, then the device executes the route
discovery process to identify the route to connect with the
destination. When the source device wants to generate a
new path to the final destination, the sender broadcast a
route request packet to the entire network and it floods in
many directions by its neighbors obtained from NT. When
the neighbors collect the RREQ it produces a reverse path
to the source device. By these way neighbor devices of next
hop to the sender established. While forwarding the RREQ
and RREP the reverse entry and forwarding entry is formed
respectively. The middle devices re-broadcast the RREQ to
their neighbors and update the device address as value[i] in
their RT as
RTTable.count++
RT += RTTable.value[i]
These message exchanges will be used to form the reverse
path for route reply RREP from the destination. The
devices collecting these packets are cached from source
and when the link is disconnected by using this it sends
RERR packet which holds information about devices that
are unable to access. After constructing the path, the
devices forward the data according to the constructed path
from source to destination.
Due to mobility, if link failure occurs then the route
reconstruction process is invoked to identify the alternate
route to connect to the destination device. During data
transmission, a packet either gets dropped or gets delayed
due to network congestion.
Figure 3. Router Congested Node
In transport layer, congestion control mechanism is
operated to identify and to control the network congestion
in base model using the dedicated parameters such as
Vol. 16, No. 4, April 2018
ISSN 1947-5500

Middle Packet Setback Distinction (MPSD) and temporary
bits per second (TB). In this Initial MPSD = 0.
Update MPSD count++
The Sum of MPSD is computed using the formula
MPSD += MPSDValue[i];
Average of MPSD is computed using the formula
AvgMPSD=
ΣMPSD
MPSDCount
D MPSDCount++
D MPSD += ( − [ ])
Computed variance of MPSD
VarianceMPSD = D MPSDCount / MPSDCount
For MPSD an upper bound (UNMPSD) and lower bound
(LBMPSD) are defined as threshold points.
StddevMPS= √
LBMPSD = AvgMPSD - StddevMPSD
UBMPSD = AvgMPSD + StddevMPSD
If the current statistics of the network crossed these
threshold boundaries then the network is marked as
congested network.
if(MPSD < LBMPSD || MPSD >UBMPSD)
Enable congestion Flag = 1
Also the TB computed as
TBCount++
TB += TBvalue[i];
AvgTB = TB / TBCount;
D TB += ( − [ ])
MBS is computed based on idle packet state; success
packet state and collision state at channel usage period and
are also based on the number of transmissions as success,
collision and idle states as given below. Channel usage
(CU) estimation provides the present utilization of overall
bandwidth consumption.
PI = pow((1-tow),n);
PS = ¬(1 − ¬ )
PC = 1 - Pidle - Psuccess;
TS = TData + TAck + SIFS + DIFS
TC = TData + Timeout + DIFS
TI = 1 - TSuc – Tcol
CU=
PSxTS
(PIxTI+PSxTS+PCxTC)
MBS =
(PSxTS + PCxTC)
(PIxTI+PSxTS+PCxTC)
Due to link failure and node failure, these parameters may
provide false deviations as the congested state of the
network. In order to avoid these false positive cases, the
metrics are extended as follows, Completion Time (CT) is
taken as
D CT += ( − [ ]) ,
Packet Loss Case (PLC) is computed as
D PLC += ( − [ ]) , and Packet Delay
(PD) along with the Medium Busy State(MBS).
PD=
ΣPD
PDCount
Vol. 16, No. 4, April 2018
ISSN 1947-5500

For these extended parameters, both lower and upper bound
are defined as threshold limits which are used to classify
the current network situation as congested or not. If it is
congested then congestion flag is enabled.
Update MPSDadd(MPSD)
Update TBAdd(TB)
Enable CongestionFlag
Each parameter is computed using the statistical model
related to the Gaussian distribution with mean and standard
deviation. To compute these parameters, the corresponding
information is maintained in the newly added header called
TCP Congestion header (CH). CH is computed as the time
difference between data packet sent and acknowledgment
received.
AvgDelay = αAvgDelay + (1-α) x CHDS
CHMBS Time = Max(CHMBSTime,MBS)
CurrenMBSTime = CH MBSTime
CurrentDelay =CHDS
Source node attaches the send time information in CH,
destination receives the information and copy it in CH of
ACK message. Once the ACK message reaches the source
end, it calculates the difference between the ACK packet
received time and Data Packet sent time as CH. MPSD is
computed as average delay. Transmission delay is
computed as the time difference between the packets sent
and packets received for both data and acknowledgment
messages.
AvgDelay = α * AvgDelay + (1-α) * CHDS
CHMBS Time = Max(CHMBSTime,MBS)
CurrenMBSTime = CH MBSTime
CurrentDelay =CHDS
If (Enclosed Reaction == 1)
CHMBSTime = ƀ * AavgMBS + (1-b) *
CurrentMBSTime
CHDS = b *AvgDelay + (1-b) *CurrentDelay
CH NewRate = NewRate
PLC is computed using the packet sequence number
difference at the destination end. It is calculated as the
ratio between the number of packets not in order and the
total number of packet transmitted. TB is computed as
average number of bits transmitted in cycle duration.
Using the current data rate, data packet size and
acknowledgment size, approximate channel occupancy
time are computed for both data packet and
acknowledgment packet. Back-off timeout period is also
computed using the channel occupancy time, propagation
delay and retransmission delay. Using the individual slot
time for each node, idle probability, transmission
successful probability and collision probability are
estimated. The successful transmission duration is
computed as the sum of Data duration, Acknowledgment
duration, SIFS and DIFS periods.
The collision duration is computed as the sum of data
duration, timeout period and DIFS periods. From the unit
slot duration, idle slot period is computed using the
difference between unit with sum of successful
transmission duration and collision duration. MBS is
computed using the ratio product of successful transmission
duration and successful transmission probability with entire
probability duration. MBS is computed as the ratio product
of successful transmission duration with collision duration
Vol. 16, No. 4, April 2018
ISSN 1947-5500

and successful transmission probability along with collision
probability to the entire probability duration as delay.
If (PacketType == ACK)
If (CongFlag == 1)
Compute QueueDelay = CURRENT_TIME - CHTime
CH Ds += QueueDelay
CHCountDs++
CHDS =CH Ds /CHCountDs
If (AveDelay == 0)
AvgDelay =CHDS
The MBS is estimated in Mac layer and attached in the CH
header field named MBS metric.
B. Fuzzy Inference System
The fuzzy inference system operates at destination
node. After collecting delay, ongoing rate and MBS
parameters, they are given as an input to the FIS based
congestion detection system
Fig.4 Fuzzy based Congestion Detection System
which employs fuzzification, rule set matching and
detection of congestion state. In fuzzification stage, using
the lower and upper bound the input parameters are
converted into fuzzy linguistic variables.
FuzzyIinput = MBS Delay RS
FuzzyRule = RuleSetMatching (Fuzzy1,Fuzzy2,Fuzzy3)
CongState = CongestionState(FuzzyRule)
Set CongState
The linear relationship growth model is used in rule set
formation since these parameters have uniform deviations
in the parametric values. Identified fuzzy linguistics are
compared with the rule set to compute the exact match in
the rule set to select current state of congestion. Congestion
states are classified as VLOW, LOW, MEDIUM, HIGH,
VHIGH.
CongestionState (FuzzyRule)
if(rule == VLOW)(0,0.2)
if(rule == LOW) (0.2,0.4)
if(rule == MEDIUM) (0.4,0.6)
if(rule == HI)(0.6,0.8)
if(rule == VHI)(0.8,1.0)
If congestion is detected in the network state, then the CH
is marked in the acknowledgment message. Once the
acknowledgment packet with congestion notification
reaches the source node, it reduces the current data rate to
decrease the outgoing packet count.
If congestion state is VLOW or LOW then destination
calculates new data rate using multiplicative increase. If
congestion state is MEDIUM then destination calculates
new data rate using additive increase. If congestion state is
HI or VHI then it calculates new data rate using
multiplicative decrease. It updates the new data rate and
sends the acknowledgement packet to sender. Intermediate
nodes increase or decrease congestion window size as per
the new data rate they receive from the neighbor.
C. E-MODEL
E-model is the common ITU-T transmission rating model.
This computational model can be used to measure the
quality of Voice data that help ensure that users will be
satisfied with end-to-end transmission performance. The
Sender Receiver
Rate Control
FIS
Intermediate Nodes
Vol. 16, No. 4, April 2018
ISSN 1947-5500

main output of the model is a scalar rating of transmission
quality.
It is the most commonly used method for predicting the
quality of voice signal on user side.
The ITU-TG.107 defines the relationship between R and
MOS as follows:
MOS = 1 for R<0
MOS = 1+0.035R+R(R-60)*(100-R) 7*0.000001 for
0<R<100
MOS = 4.5 for R>=100
The rating factor R is composed of the basic formula:
R = Ro – Is – Id – Ie-eff + A
Ro (signal to noise ratio) is a mathematical summary of
how the voice levels compare to the different noise
sources including circuit noise and room noise.
Id (delay impairments) is a mathematical summary of
transmission delay, talker echo and sidetone.
Is (simultaneous impairments) considers non-optimum
sidetone, quantizing distortion, overall loudness and
other impairments which occur more or less simultaneously
with the voice transmission.
Ie (equipment impairment) and A (Advantage Factor) are
both single value quantities.
To assist with calculations, default values and permitted
ranges have been established.
This calculates MOS value based on the given criteria[3].
The following table shows the relationship between R
factor and user satisfaction.
R Factor MOS User Satisfaction
140 4.5 Very Satisfied
80 4.024 Satisfied
IV. RESULT ANALYSIS
Simulation compares the proposed approach CLHCC with
FARCC and ADTCP by varying the number of nodes and
simulation time. Results show that the performance of the
proposed approach at network level and at user level. It
outperforms well than the two approaches in terms of
packet delivery ratio, end-end-delay, through put. It
calculates user level quality called Quality of Experience
(QoE) using E-Model. MOS values show that the proposed
approach attains user dissatisfaction level in terms of
quality.
Fig 5 : Node Versus Packet Delivery Ratio
Figure 5 shows the comparison between three approaches
based on the metric Packet Delivery Ratio. The graph
clearly indicates that the proposed approach outperforms
well than the existing approaches.
85.0000
87.0000
89.0000
91.0000
93.0000
95.0000
97.0000
99.0000
101.0000
50 75 100 125 150
Pkt_Delivery_Ratio
Node
Pkt_Delivery_Ratio
Pkt_Delivery_Ratio
_Base
Pkt_Delivery_Ratio
_Adtcp
Pkt_Delivery_Ratio
_CLHCC
Vol. 16, No. 4, April 2018
ISSN 1947-5500

Fig 6 : Node Versus End-End-Delay
Figure 6 shows the performance of proposed approach in
terms of End-End-Delay. It suffers from less End-End-
Delay than the existing approaches.
Fig 7 : Node Vs Throughput
Figure 7 shows the maximum through put attainment of
proposed approach.
Fig: Node Versus MOS
Figure 7 shows that user level quality of Voice. Quality
remains at same level as the performance increases.
V. CONCLUSION
In Mobile Ad-hoc Network, MBS, TB, Delay distinction
are used to measure the congestion level of the network.
These investigations support us to change transmission
rate. In this paper, we checked congestion level in MANET
with Fuzzy Inference System to finalize the data rate and to
control the flow of data at sender by the receiver
dynamically. It also adjusts the congestion window as per
the new data rate on the stamp field. Hence it supports both
proactive and reactive approach to congestion detection and
to control congestion. The proposed approach outperforms
well for network level quality and user level quality than
the existing congestion control approaches ADTCP and
FARCC.
REFERENCES
[1] Gavin Holland and Nitin Vaidya, “Analysis of TCP Performance over
Mobile AdHocNetworks”, ACM/IEEE International Conference on
Mobile Computing and Networking (MobiCom '99), (Seattle,
Washington), August 1999.
0.00000
0.10000
0.20000
0.30000
0.40000
0.50000
0.60000
0.70000
50 75 100 125 150
End-End-Delay
Node
End-End-Delay
Delay_Base
Delay_Adtcp
Delay_CLHCC
0
20
40
60
80
100
120
50 70 90 100 120
T
h
r
o
u
g
h
p
t
Node
Throughput Through put Base
Through put Ad-TCP
Throughput CLHCC
4
4.1
4.2
4.3
4.4
4.5
4.6
0 100 200 300
MOS
Node
Node Vs MOS
MOS Value
CLHCC
Vol. 16, No. 4, April 2018
ISSN 1947-5500

[2] Jane y. yu and Peter h. j. Chong,“A Survey of Clustering
Schemes for Mobile ad hoc Networks”,IEEE Communications
Surveys &Tutorials, First Quarter, 7,
2005.
[3] Lingfen Sun, Emmanuel C. Ifeachor, “Voice Quality
Prediction Models and Their Application in VoIP
Networks”, IEEE Transactions On Multimedia, Vol. 8, No.
4, August 2006.
[4] Foez ahmed, Sateesh Kumar Pradhan, Nayeema Islam, and
Sumon Kumar Debnath, “Performance Evaluation of TCP
over Mobile Ad-hoc Networks”, International Journal of
Computer Science and Information Security, Vol. 7, No. 1,
2010.
[5] Praveen Dalal,“Study on Transport Layer Protocols for
Wireless Ad-Hoc Network”,Proceedings of the 5th National
Conference;INDIACom Computing For Nation
Development, March 10, 2011.
[6] Sreenivasa B.C,G.C. Bhanu Prakash, K.V.Ramakrishnan,
“Comparative analysis of ADTCP and
MADTCP:Congestion Control Techniques for improving
TCP performance over Ad-hoc Networks”,International
Journal of Mobile Network Communications &
Telematics (IJMNCT) Vol.2, No.4, August 2012.
[7] Abdelhak Bentaleb, Abdelhak Boubetra, Saad
Harous,“Survey o Clustering Schemes in Mobile Ad hoc
Networks”, Scientific Research Communications and
Network, May 2013.
[8] H. Zare, F. Adibnia, V. Derhami,”A Rate based Congestion
Control Mechanism using Fuzzy Controller in MANETs”,
International Journal of Computer Communication, June
2013.
[9] GargiParashar, Manisha Sharma,“Congestion Control in Manets
Using Hybrid Routing Protocol”, IOSR Journal of Electronics and
Communication Engineering,Vol.6, Issue 3, Jun 2013.
[10] Bandana Bhatia, Neha Sood, “AODV based Congestion Control
Protocols: Review”, International Journal of Computer Science
and Information Technologies, Vol.5(3) , 2014.
[11] Waleed S. Alnumay,“Security Enhanced Adaptive TCP for
Wireless Ad Hoc Networks”, Journal of Information Security, 2014.
[12]Said El brak, Mohamed El brak, Driss Benhaddou “A New QoS
Management Scheme for VoIP Application over Wireless Ad Hoc
Networks”, Journal of Computer Networks and Communications, Vol.
2014.
[13] Abinasha Mohan Borah, Bobby Sharma, Manab Mohan Borah,“A
Congestion Control Algorithm for Mobility Model in Mobile Ad-hoc
Networks”, International Journal of Computer Applications, Volume
118 – No.23, May 2015.
[14] Kaushika Patel, Nayana Ram, Virendra Barot,“TCP in MANET :
Challenges and Solutions”, International Journal of Innovative
Research in Science, Engineering and Technology,Vol. 4, Issue 12,
December 2015.
[15] Ritika Mehra and Manjula Saluja, “Adaptive Congestion
Control Mechanisms in Mobile Ad-Hoc Networks”, International
Journal of Engineering Development and Research,
Volume 5, Issue 1,2017.
Vol. 16, No. 4, April 2018
ISSN 1947-5500

Cross Layer Based Hybrid Fuzzy Ad-Hoc Rate Based Congestion Control (CLHCC) Approach for VoMAN to Improve Quality of VoIP Flows

More Related Content

What's hot (17)

Similar to Cross Layer Based Hybrid Fuzzy Ad-Hoc Rate Based Congestion Control (CLHCC) Approach for VoMAN to Improve Quality of VoIP Flows (20)

Recently uploaded (20)

Cross Layer Based Hybrid Fuzzy Ad-Hoc Rate Based Congestion Control (CLHCC) Approach for VoMAN to Improve Quality of VoIP Flows